What IFRA Took
A complete guide to the restricted list, the science behind it, the creative cost of it, and the case that a rare and manageable risk has been answered by abolishing the materials instead of labelling them.
Somewhere in a formula written before the First World War there is a line that reads oakmoss absolute, and beside it a number a modern perfumer would not be permitted to write. The lichen has not changed. The chemistry has not changed. What changed is a document.
That document is the IFRA Standards Library, published by the International Fragrance Association, and it is the most consequential piece of writing in modern perfumery. It is not law. No parliament voted for it. No inspector enforces it. And yet almost every fine fragrance on a department store counter today has been shaped, trimmed, or rebuilt around it.
This guide is not neutral, and it should say so at the start. It sets out what the Standards actually say, material by material, with the numbers attached, and it takes the counterarguments seriously enough to print them in full. Having done that, it reaches a conclusion: the settlement is disproportionate. A condition affecting around two per cent of adults, unpleasant but rarely serious and almost entirely avoidable once it is known about, has been answered by removing materials from everybody, permanently, on the authority of a private association whose science is funded by the companies that sell the substitutes.
Nobody serious disputes that fragrance allergy exists. The dispute is why a rare and manageable risk was met with abolition rather than a label.
- IFRA is a trade association, not a regulator. Its seven Regular Members make roughly four fifths of the world's fragrance, and its Standards bind nobody in law. Compliance is commercial, achieved through the supply chain rather than through any statute.
- The restrictions are severe and specific. Oakmoss at 0.10%, rose ketones at 0.043% summed, methyl eugenol at 0.011%, safrole at 0.01%, plus outright prohibitions. Every one of those is a Category 4 figure, taken from the most permissive skin contact category there is.
- What was removed was not decoration. It was friction. The damp, bitter, animal, tarry and burnt registers were the things that gave a composition a floor and an edge. A palette can lose beauty and still function. Losing its roughness is structural.
- The science is paid for by the industry it assesses. RIFM reported revenue of $12,093,553 in 2024, none of it in contributions, on dues scaled to members' sales. Three named individuals sit on both the IFRA and RIFM boards. The underlying database is behind a paywall.
- The companies writing the rules sell the replacements. Every restricted natural has a branded synthetic from a house at the table, and a patented molecule earns for twenty years where a commodity natural earns nothing.
- Forty years of tightening has not moved the curve. A 2024 meta-analysis of eighty four studies published between 1981 and 2022 reports its finding on the trend line in one sentence: no clear time trends were observed.
- Nobody audits any of it. The Certificate of Conformity is written by the fragrance supplier. IFRA states plainly that it neither creates nor endorses those certificates, and it publishes no routine audit or analytical verification programme.
- Most of the world never signed up. The United States restricts none of these materials. India, the Gulf and the whole of Southeast Asia require no fragrance allergen labelling at all. The regime is a European settlement wearing an international name.
Perfumery, a hundred-year-old art, has taken a long time dying, but on January 1, 2010 it will be officially dead.Luca Turin, NZZ Folio, April 2009
The machinery
IFRA is not a regulator. That is the first thing to understand, and the most contested.
The International Fragrance Association was founded in 1973 as a trade body. Its Regular Members are the seven largest fragrance suppliers in the world: dsm-firmenich, Givaudan, IFF, Symrise, BASF, Robertet and Takasago. Between them they account for roughly four fifths of global fragrance production by volume.1
Its Standards are binding on its members through the IFRA Code of Practice. On everyone else they are binding through commerce alone. A brand that wants a retailer, a contract manufacturer or a liability lawyer to take it seriously will be asked for an IFRA conformity certificate, and so an estimated eighty to ninety per cent of the industry complies with a rulebook that has no statutory force whatsoever.2
The word does a great deal of work. It is accurate in law and misleading in practice. Nothing compels a perfumer to comply. Almost nothing permits them to sell at scale if they do not.
The science comes from a second organisation. The Research Institute for Fragrance Materials, founded in 1966, gathers and generates the toxicological data. An Expert Panel of external academics reviews that data and issues conclusions. IFRA then converts those conclusions into Standards. The relationship between these two bodies is the whole of chapter nine, and it is where the argument gets uncomfortable.
Hold on to one structural fact from the outset, because chapter thirteen returns to it. This is a system with real consequences and no inspection. IFRA's own guidance states that the Certificate of Conformity is a document established by the companies creating fragrance mixtures, that IFRA does not create the certificates, and that no company issues them on IFRA's behalf. Conformity is declared by the seller of the material. The rulebook constrains the people who follow it and never once checks the bottle.27
Every number on the restricted list is a division problem.
Since the 49th Amendment the method has been called QRA2, the second generation Quantitative Risk Assessment. It begins with a figure called the NESIL, the No Expected Sensitisation Induction Level, which is the dose per unit of skin at which a material is not expected to induce allergy. It is derived from human repeat insult patch testing and from animal or in vitro assay data, and it is expressed in micrograms per square centimetre.
That figure is then divided by a set of Sensitisation Assessment Factors, which account for variation between individuals, for the way a product is actually used, and for the effect of the base the material sits in. What comes out is the Acceptable Exposure Level.
The AEL is then compared with the CEL, the Consumer Exposure Level, calculated using the Creme-RIFM aggregate exposure model. If exposure exceeds the acceptable level, the permitted concentration is cut until it does not.
Notice what that chain of divisions does. Every uncertainty factor is a judgment call, every one is applied in the same direction, and each one is individually defensible. Stack enough conservative judgments on top of one another and you arrive at a number that no single decision produced and nobody has to defend as a whole.
The 49th Amendment then changed one thing that matters more than any single material restriction. Exposure is now aggregated. The model no longer asks how much of a substance you meet in one perfume. It asks how much you meet across every scented product you might use in a day: the shampoo, the deodorant, the body lotion, the laundry, the fabric softener, the candle, and then the perfume. Fine fragrance is a small fraction of that total and absorbs a proportionate share of the cut. The art is billed for the detergent.
Products are sorted into twelve categories, raised from eleven in 2020. Category 1 covers products applied to the lips. Category 5 covers face and body creams. Category 4 is fine fragrance, and IFRA's own guidance defines it as hydroalcoholic and nonhydroalcoholic fine fragrance of all types, which expressly includes solid perfume and fragrancing cream. It is the most permissive of the skin contact categories. Every figure in the ledger below is a Category 4 figure. It is the best case.3
One detail in that guidance decides a great deal later in this guide. A limit is expressed as a maximum concentration in the finished consumer product, not in the fragrance concentrate. An eau de parfum at twenty per cent concentrate can therefore carry a restricted material at five times its Category 4 number inside the concentrate itself. An oil that is sold and worn neat cannot, because the oil is the finished product.
Below all of this sits a floor. The Dermal Sensitisation Threshold sets a level under which no material specific data is required at all: 900 micrograms per square centimetre for chemicals judged non-reactive, and 64 for reactive ones. It is a pragmatic device for the very large number of materials nobody has money to test individually.
The ledger of losses
The empty compartment is the argument in miniature. A restriction removes a material, but the space it occupied in a formula remains structural.
What is restricted, and by how much.
The library runs to more than three hundred Standards. What follows is the part a perfumer actually feels: the materials whose restriction changed the sound of the instrument. Every percentage is the maximum permitted in the finished consumer product in Category 4, taken from the individual Standards of the 51st Amendment, in force since 30 June 2023.4
Limonene and linalool are not capped by concentration. Their Standards are peroxide specifications: the material must be kept below twenty millimoles of peroxide per litre. The sensitiser is the oxidation product, not the molecule, so the rule governs handling and storage rather than dosage.
Musk ketone is not prohibited. It carries a purity specification limiting residual musk xylene. It remains usable, and is used.
Civet, castoreum and natural deer musk have no IFRA Standard at all. Their disappearance is a matter of sourcing, cost, ethics and the Convention on International Trade in Endangered Species. Attributing it to IFRA is one of the most common errors in the debate, and it damages the arguments that are correct.
The chypre was built on a material that is now one part in a thousand.
Coty's Chypre of 1917 gave its name to a structure rather than a perfume: bergamot at the top, labdanum in the middle, oakmoss underneath. The lichen does two jobs at once. It supplies a specific smell, inky and wet and bitter, closer to forest floor than to any flower, and it acts as the fixative that carries the volatile citrus down into the base so that the composition reads as one continuous object rather than two.
Atranol and chloroatranol, two constituents of the lichen, were identified as unusually potent sensitisers. IFRA responded with a limit of a tenth of one per cent conditional on those constituents being held below a hundred parts per million. In the European Union the atranols went further and were prohibited outright by Regulation (EU) 2017/1410, with total withdrawal from the market on 23 August 2021.
Oakmoss in the formula
At the current limit the base has no floor. The citrus evaporates on its own schedule, the labdanum sweetens without a counterweight, and the composition reads as two perfumes played in sequence rather than one structure.
There is a further turn of the screw the debate usually misses. The oakmoss a perfumer can buy today has already been through fractional distillation to strip the offending constituents out. The critic Kafkaesque has argued the point forcefully: the commercial material is already compliant before anybody weighs it. The perfumer is not choosing between a legal dose and an illegal one. The perfumer is choosing between a small dose of an already diminished thing and nothing at all.
The honest version of the complaint is not that beautiful chypres are impossible. Several excellent ones have been made since 2008. It is that a specific texture is no longer available at the strength that made it a texture, and that the substitutes solve the smell without solving the structure. Evernyl is a fine material and a poor fixative. What was removed was not only a note. It was a load bearing wall.
When one term changes, every relationship around it changes too. The work is not substitution. It is rebuilding the architecture.
If we ban citral from perfumes, of which certain elements are allergens, we should ban orange juice. It is absurd.Frederic Malle
Four moments that changed how perfume smells.
It helps to be precise about dates, because the debate is full of confident claims attached to the wrong amendment. The library has been revised more than fifty times. Most revisions are small. Four are not.
The open era
IFRA is founded and begins publishing Standards, but the library is short and the great naturals are largely untouched. A perfumer of this period can dose oakmoss, nitro musks and animalics at levels that would now be unthinkable. Everything the connoisseur means by the phrase they do not smell like that any more was composed under these conditions.
Between those four moments sit the smaller amendments that nonetheless matter: the 47th in 2013, which prohibited crude birch tar and crude cade and took the smoked leather register with them, and the 50th in 2021, whose entire content was a single prohibition on mintlactone. The 52nd Amendment went to consultation from 12 December 2025 to 12 June 2026, with notification expected at the end of November 2026. Nothing about this system has finished happening.
The rules that actually carry legal force are written somewhere else.
Perfumers blame IFRA. In the European Union it is Regulation (EC) No 1223/2009 that has teeth. Annex II lists what is prohibited, Annex III what is restricted. Every product needs a Responsible Person established in the Union, a notification on the CPNP portal, and a Cosmetic Product Safety Report signed by a qualified assessor. Article 37 leaves penalties to national law, so enforcement sits with member state authorities and ranges from corrective action to withdrawal, recall and fines.18
Two of the most disruptive removals of the past decade came from Brussels and not from Geneva. HICC, together with atranol and chloroatranol, was moved to Annex II by Regulation (EU) 2017/1410, with no new products from 23 August 2019 and total withdrawal from 23 August 2021.19 BMHCA, sold as Lilial, was classified as a reproductive toxicant and prohibited from 1 March 2022 by Regulation (EU) 2021/1902, while its IFRA Standard still sits at 1.4%.5
The third instrument is the one currently reshaping the industry, and it prohibits nothing at all. Regulation (EU) 2023/1545, adopted on 26 July 2023, expands the individually declarable fragrance allergens to eighty one substances, named on the label above 0.001% in a leave-on product and 0.01% in a rinse-off. Among the new entries are lavender oil, rose oil, jasmine, eucalyptus, lemongrass and sandalwood.6
Annex II
Prohibited. Nothing on this list may be present in a cosmetic product placed on the European market, at any concentration, in any category.
Annex III
Restricted. Permitted under stated conditions of concentration, product type, purity or labelling. This is where most fragrance materials live.
The declaration, and why it matters more than a ban
A brand facing an ingredient list with thirty allergen names has a commercial reason to reformulate toward materials it does not have to declare. That reason is not safety. It is the appearance of safety. Regulation reshapes taste most effectively not by forbidding beauty but by making some beauty expensive to admit to.
31 July 2026 is the last date non-compliant product may be placed on the European market. 31 July 2028 is the last date it may be made available to a consumer.
One nuance is worth holding on to, because chapter twelve turns on it. A Cosmetic Product Safety Report can lawfully pass a formula that exceeds an IFRA Standard, provided the assessor substantiates the safety of that dose and no Annex II or Annex III limit is breached. IFRA conformity is the ordinary route to substantiation. It is not the only one, and it is not itself the law.
What perfumery can no longer say at volume.
Set aside the arguments for a moment and describe the sensory result, which is the part that rarely gets written down carefully. Three registers have thinned.
The floor
Oakmoss and treemoss at a tenth of one per cent, crude birch tar and crude cade gone, styrax and opoponax in trace. What these shared was not a smell but a function: they put something underneath a composition that was damp, bitter, mineral and slightly unpleasant. A perfume with a floor feels like a place. A perfume without one feels like a bouquet held at arm's length. This is the loss perfumers mean when they say modern fragrance smells like it is hovering.
The animal
The nitro musks are gone, and the polycyclic and macrocyclic musks that replaced them are cleaner, whiter and more even. Costus is prohibited. Civet and castoreum declined for reasons of ethics and supply rather than regulation, but they declined. The base register of contemporary perfumery is more comfortable and less specific, and it is very difficult now to make a perfume smell like a person rather than like a fabric.
The burn
Safrole at one hundredth of a per cent, methyl eugenol at eleven thousandths, isoeugenol at eleven hundredths, cinnamal at a quarter, rose ketones at forty three thousandths summed. The spice and the jam and the smoke all narrow at once. What survives is sweetness without heat, which is the defining character of the mainstream oriental of the past fifteen years.
What these materials have in common is not that they smelled good. Several of them, alone, smell frankly bad. It is that they provided friction, and friction is what gives a composition the impression of having an edge, a floor and a body. A palette can lose a great deal of beauty and still function. Losing its roughness is a structural injury.
And the difficulty of reformulation is not that a substitute cannot be found. It is that a formula is a system of relationships, and changing one term changes all of them.
There's no such thing as getting a password nearly right. If there's one letter wrong, the thing doesn't click.Luca Turin, on reformulation
The argument
Twenty six voices on the record.
Most writing on this subject quotes the same two or three lines. What follows is the fullest collection we could assemble of people who actually make perfume, or govern the rules, speaking for attribution. Where a quote reached print second hand, through a wire agency relayed by a blog, it is labelled as such. Use the filter to read only one side.
Perfumery, a hundred-year-old art, has taken a long time dying, but on January 1, 2010 it will be officially dead.
The absurd IFRA regulations have nixed many staple materials. Bear in mind the principle of IFRA is so daft that it could only have been dreamt up with the help of dermatologists. That is how the perfumer's palette shrank, and that is why very powerful materials now rule.
NZZ Folio, April 2009, and Substack, 13 April 2024. Source
The decisions surrounding ingredients at IFRA and their RIFM and Prodarom colleagues are not based solely on science.
Screened by a non-diverse panel of self-recruited scientists.
BeautyMatter, 26 June 2022. Source
Who are the members of IFRA? You will see that the big industry is in there, as members, like IFF. Thus, all regulations are basically influenced by the big industry, too.
The restrictions imposed by EU will kill many suppliers of essential oils and absolutes. I am faced with a narrowing market for high quality essential oils.
Perfume Shrine, 6 April 2009. Source
Due to the IFRA regulations, there are some materials that I can't even purchase anymore.
If you make a fractional distillation and you pull out what the European Commission doesn't want any more, then you create an olfactive hole.
Basenotes, 4 November 2011, and Persolaise, August 2014. Source
If we ban citral from perfumes, of which certain elements are allergens, we should ban orange juice. It is absurd.
It can take more than six months to reformulate a perfume, and a minimum of some thirty tests. So to protect a small portion of the population, we are making the rest suffer.
Reported in Kafkaesque's series on the industry and EU regulation, 2014. Source
The regulations drive me completely insane and are yet another hoop to jump through, more expense and paperwork which basically give the consumer a list of ingredients which are actually allergens and not ingredients.
If I had to reduce the amounts of jasmine, clove, tolu and rose in Anubis it simply wouldn't be Anubis anymore.
Kafkaesque, 5 February 2015. Three fragrances generated a ninety six page safety report. Source
The recent proposals made by the European Commission, if they become law, threaten to destroy fine perfumery as we know it.
Reformulation is a scientistic utopia that denies the unique emotion created by the alchemy of assembled molecules.
Open letter to the EU public consultation, March 2014. Second quote translated from the French. Source
Big brands tell me: replace this and that and make sure it smells the same and costs the same to produce.
Once you change an ingredient or two it can be very difficult to keep the scent absolutely intact, especially if those ingredients played an important role in defining the scent.
Reuters, carried by The Business of Fashion, 2014. Source
I expected big groups to take the initiative on this matter but it turns out that they are the most risk averse.
Grain de Musc, on IFRA and reformulation. Source
Eau Sauvage was a real chef d'oeuvre in its original form. It used to be very green and fresh. Today, it has been replaced by something softer and duller.
Reuters, relayed by Kafkaesque, 1 January 2013. Second hand relay. Source
I know the original formula by heart and I imagine they had to change progressively because of new IFRA regulation.
Reuters, relayed by Kafkaesque, 1 January 2013. Second hand relay. Source
Many perfumes have had to be reformulated even though they were considered masterpieces due to changing legislation.
He compared it to changing the colours of the Mona Lisa. Reuters via BoF, 2014. Source
I think Brussels' focus is a little exaggerated specially compared to alcohol and cigarettes which are sold freely and do more harm than perfume.
Reuters, carried by The Business of Fashion, 2014. Source
Bureaucracy Destroys Art!
Entire genres of classic fragrances are being either discontinued or completely reformulated.
Washington Post Magazine, 21 April 2021, and Luca Turin, 2024. Source
I'm thinking that it isn't so much the formula they're protecting as the story.
Washington Post Magazine, 21 April 2021. Source
The whole industry is smoke and mirrors. It's all air. Literally air.
Washington Post Magazine, 21 April 2021. Source
Give the people the strength to really choose what they want. If you tell them a story, they won't trust you anymore.
On disclosure as the alternative to restriction. Washington Post Magazine, 2021. Source
The Guild joins Cropwatch in asking for a moratorium on the IFRA 40th amendment.
The only organised, named dissent on record, 2007. Source
It is best to prioritise regulatory compliance over every other aspect, even odour, when designing new aroma molecules.
To the British Society of Perfumers, via Basenotes. Not a complaint, but evidence of what the regime did to the culture of the trade. Source
There's never been a question of ban, anyway, it is about potential restrictions.
One aspect is labelling, because people have allergies and it's good for people to know if they are allergic to something.
Basenotes, 28 May 2013. He deflates the panic rather than joining it. Source
Yes, but as long as one can conform to the new rules, it will hopefully be ok.
Answering whether French perfumery is in danger. Kafkaesque, 6 September 2013. Source
At Chanel, we follow very closely talks about regulation and scientific findings concerning raw materials.
Reuters via BoF, 2014. An illustration of how carefully the largest houses say nothing. Source
If in future, because of IFRA, we have to change, then we will do it, with a lot of respect.
Routinely conscripted into the case against IFRA, where he does not belong. Source
IFRA is not a lawmaker, IFRA is not even a formal regulator. It is a self-regulatory trade body.
Some brands also ask their fragrance houses or create their own reformulation for commercial reasons and blame IFRA for the change in odour.
The most detailed insider rebuttal available. Source
We broadly welcome the proposed measures.
Most perfumes which are twenty years old or more will have already been reformulated several times because science has evolved and we want to ensure the safety of consumers.
February 2014, Reuters, relayed by Kafkaesque. Second hand relay. Source
IFRA regularly introduces updates to its product stewardship program, the IFRA Standards, to ensure that consumers can continue to enjoy fragranced consumer products with confidence.
IFRA press release on the 51st Amendment, 24 July 2023. Source
I am honored to take on the role of President of IFRA and continue the organization's commitment to safe and sustainable fragrance.
Appointment release, 18 November 2024. We searched for any statement from IFRA's current president answering the criticism in this guide and found none. Source
Showing 26 of 26
A widely shared 2014 essay stated that IFRA "is not an independent entity, it is financed by Givaudan, Symrise and US-based International Flavors and Fragrances." We could not verify that financing structure from any primary document and we are not repeating it as fact. What is documented, and set out next, is different and in some ways more troubling.
The decisions surrounding ingredients at IFRA and their RIFM and Prodarom colleagues are not based solely on science.Christophe Laudamiel, BeautyMatter, 2022
Who benefits from the restricted list?
Here is the charge, stated in the terms perfumers actually use. The companies that write the Standards are the companies that sell the replacements. What follows is the evidence, taken entirely from the organisations' own published material, and then the part of the claim the evidence does not reach.
Begin with the overlap, because it is public record and not inference. Six of the largest suppliers hold seats on the IFRA Board. Three individuals sit simultaneously on the IFRA Board and on the board of the Research Institute for Fragrance Materials: Jeremy Compton of Givaudan, Catherine Gadras of Robertet, and Bhuvana Nageshwaran of Ultra International. IFRA's president Alexander Mohr and IFRA's Matthias Vey sit on RIFM's Advisory Committee, which advises on all of RIFM's scientific programmes.7
IFRA writes the Standards and enforces them through the Code of Practice.
RIFM gathers the toxicology, funds the studies, and holds the database.
The Expert Panel reviews the data and issues the conclusions.
Three named individuals sit on the boards of both IFRA and RIFM at the same time.
The seven suppliers that fund both organisations also manufacture and sell the molecules that replace what the Standards restrict.
Then the money. RIFM is funded by member dues scaled to sales, running from five thousand dollars at the lowest band to more than one million four hundred and twenty five thousand at the highest. Its 2024 return to the United States tax authority reports total revenue of $12,093,553, of which contributions were zero and programme service revenue was 95.6%. It is funded by the industry it evaluates, entirely and by design. The database holding the underlying toxicological data is available by paid subscription, with no publicly stated price.8
Then the Expert Panel, the strongest part of the system and still not clean. Its members are genuine external academics who select their own successors and operate under a written conflict of interest policy. They are also paid four thousand dollars per meeting day plus two hundred dollars per hour of preparation, and that policy expressly permits industry funding to be routed to a member's institution. Independence here is a procedural claim, carefully constructed and worth taking seriously. It is not a financial fact.9
And then the part perfumers point at, which is simply a catalogue.
A captive is a molecule a house owns and will not sell to competitors. It is the most profitable object in the fragrance trade, and it is protected for twenty years. Luca Turin has made the structural version of this argument more sharply than anyone: a company holding synthetic patents has no commercial reason to fight a restriction that pushes formulation toward patented molecules, and every reason to be relaxed about one.
There is no public evidence that any supplier lobbied for any specific restriction in order to sell a replacement. The overlap of boards is documented. The dues structure is documented. The replacement catalogue is documented. The motive is inferred, and inference is not evidence.
Two further claims circulate widely and should not be repeated: that IFRA is financed by a levy on restricted materials, and that named houses have closed because of the Standards. Neither has a documented case behind it. The argument against the current settlement is strong enough without them.
The best case for IFRA, stated fairly, and why it does not survive.
The most detailed rebuttal comes from inside the profession rather than from the association. Pia Long, a perfumer and co-founder of Olfiction, begins with a point of order most of the debate ignores: "IFRA is not a lawmaker, IFRA is not even a formal regulator. It is a self-regulatory trade body." Her argument is that without a single industry standard, national regulators would legislate separately and far more bluntly, and that the association is the dam holding back a flood that would otherwise arrive as thirty incompatible national statutes.10
Her second point is the one critics should sit with longest. "Some brands also ask their fragrance houses or create their own reformulation for commercial reasons and blame IFRA for the change in odour." A reformulation is cheap to blame on a regulator and expensive to admit as a cost decision. Nobody issues a press release announcing that the concentrate budget was halved.
Nor is opposition universal among the perfumers usually cited as authorities. Jean-Claude Ellena has said that if the rules require change then the change will be made, and made with respect. Christopher Sheldrake spent an interview deflating the panic rather than joining it. Both names appear constantly in the case against IFRA and neither belongs there.
The charge
A trade body with no democratic mandate has removed materials that took three centuries to learn how to use, on the basis of data it owns, paid for by the companies that profit from the substitutes, reviewed by a panel it pays, and published behind a subscription.
Forty years of tightening has not moved the sensitisation rate in the tested population. If the remedy worked, the curve would bend.
The defence
Fragrance contact allergy is a real disease affecting a real population, and it is lifelong. The alternative to a single industry standard is not freedom. It is thirty national regulators legislating independently, with no shared method and no appeal.
Much of what is blamed on regulation is cost, supply, or a creative decision nobody wants to sign.
Two famous cases prove that last point on their own. The cumin that vanished from Rochas Femme was removed in 1989 by Olivier Cresp as a creative revision, years before the restrictions usually blamed for it. Chanel No 19 changed character after the Iranian revolution of 1979 altered its galbanum supply, not after any Standard. And Victoria Frolova, writing as Bois de Jasmin, has argued that the reworkings of Chamade and Nahema in 2011 and 2013 improved on the versions immediately preceding them.
All of that is true, and none of it answers the objection. Take the defence at its strongest and it establishes three things: that fragrance allergy is real, that a single standard is administratively better than thirty, and that regulation is blamed for sins it did not commit. Grant every one of them. What remains unanswered is why the response to a rare and avoidable condition had to be the permanent removal of materials from everybody, decided privately, on evidence the public cannot read, by an association funded by the sellers of the alternatives.
The dam argument in particular proves less than it appears to. A dam is justified by the flood it holds back, and nobody has shown the flood. There is no counterfactual in which the European Commission, absent IFRA, would have banned oakmoss outright. What the association actually supplies is not protection from stricter law but a ready made technical basis that regulators can adopt without doing the work themselves. It is not standing between the industry and the legislator. It is drafting for them.
What the epidemiology actually shows, in both directions.
Both sides of this argument cite numbers and neither side cites all of them. Here are the three that matter most.
Clinically relevant fragrance contact allergy in the general adult population, and 0.8% on the stricter criterion, in the largest study of its kind. Diepgen et al, British Journal of Dermatology, 2015.11
Pooled prevalence of a positive reaction to Fragrance Mix I across eighty four studies published between 1981 and 2022, with 3.64% for Fragrance Mix II. Botvid et al, Contact Dermatitis, 2024.12
The conclusion of that same review on the direction of travel across four decades. Restriction tightened continuously through the period. The sensitisation rate in the tested population did not visibly fall.
The case for restriction is not weak. The Scientific Committee on Consumer Safety, in opinion SCCS/1459/11, identified eighty two established contact allergens in fragrance and estimated fragrance contact allergy at one to three per cent of the general population, rising to roughly sixteen per cent among patients already presenting with eczema. This is a genuine condition affecting millions of people, and it does not resolve.13
The careful reading of the Botvid finding is also necessary. Patch tested populations are selected for suspected allergy and are not the general public. A stable rate in a market that has grown enormously and become far more scented over the same forty years is not the same as a rate that has not responded.
But the burden of proof has moved. A regime that has removed this much from an art form, over this long a period, at this cost, ought by now to be able to demonstrate what it purchased. That demonstration has not been produced, and the data that exists does not obviously supply it.
The world outside the jurisdiction
A very large part of perfumery never complied, and nothing happened.
Everything above describes a European settlement wearing an international name. IFRA is global in membership and its Standards are written for the world, but the only place they are reinforced by hard law is the European Union, and even there the law is a separate instrument with a shorter list. Step outside that jurisdiction and the picture changes completely.
It is worth being exact about what actually happens to a house that ignores the Standards, because the folklore is much worse than the reality. No regulator anywhere can act against IFRA non-compliance as such. Conformity is a private contractual matter between a fragrance supplier and its customer. What can happen, in ascending order: the house cannot sell into the European Union or the United Kingdom without a Responsible Person and a notification; marketplaces can delist it for want of a safety report or a complete ingredient list; wholesale and department store accounts will ask for a conformity certificate and decline without one; and if a customer is actually injured, non-conformity becomes evidence that the maker fell below the industry standard of care, in an ordinary product liability claim, rather than a breach of any statute.
That is the whole of it. Not a fine, not a prosecution, not a recall. A commercial door closes, and it is a door most of the world's perfumers were never going to walk through anyway. One claim in wide circulation should also be retired: it is often said that insurers void product liability cover for non-compliant perfume. We could not source that to any policy document and we are not stating it as fact.
The ASEAN Guidelines for the Safety Assessment of a Cosmetic Product ask that a safety evaluation take account of "conformity to the latest International Fragrance Association (IFRA) Guidelines which have to be provided by the supplier," and list a "Certificate of compliance with latest IFRA Guidelines" among the contents of the Product Information File.
Read that carefully. Across ten countries and roughly seven hundred million people, IFRA's function is to have supplied a piece of paper. The binding list is the ASEAN annex. The certificate is a file item. Nothing in the system compares the certificate with the liquid.
Indonesia is the clearest case, and the most instructive, because it is not a peripheral market. Perfume there is a cosmetic requiring notification to BPOM, filed by an Indonesian legal person, valid three years, checked against ingredient annexes rather than against a toxicological review of the formula. Lilial was written into Indonesian black letter law only in October 2025, with a twelve month transition, three and a half years after the European ban. From October 2026 halal certification becomes mandatory for cosmetics under Law 33 of 2014, which for perfume turns on the provenance of the ethanol rather than its percentage, following the Indonesian Ulema Council fatwa of 2018.
What enforcement actually looks like is more revealing still. In the first half of 2026 BPOM's intensified operation seized illegal cosmetics worth 35.8 billion rupiah, and 86.8 per cent of the findings were products with no distribution permit at all, rather than products containing a prohibited ingredient.24 The system polices registration. It does not police composition. That is not an Indonesian failing. It is what every notification based regime does, including the European one, and it is worth remembering before anyone describes IFRA compliance as a safety net.
And then the fact that ought to end the conversation about who these rules are for. Indonesia supplies roughly ninety per cent of internationally traded patchouli oil, twelve to fifteen hundred tonnes a year. It produces around sixty two per cent of the world's cloves and about eighty per cent of its nutmeg. Garut in West Java grows vetiver. Sumatra produces the benzoin that has scented European churches for four centuries.28 The country that grows the base of modern perfumery has no seat at the table where the base is regulated, no national association, and no vote on the amendment that decides what its harvest may be used for.
Meanwhile its own market is among the fastest growing in the world. Indonesian fragrance e-commerce was worth roughly one billion dollars in 2025 and is projected near two billion by 2029, with local brands taking the top places on the domestic marketplaces.29 None of that growth is happening under an allergen declaration requirement, and the sky has not fallen.
An IFRA limit is a percentage of the finished consumer product. An eau de parfum at twenty per cent concentrate can carry a material at five times its Category 4 number inside the concentrate. An attar is the finished product, so the number applies to the bottle itself.
That is not a loophole in the attar's favour. It is the opposite. Category 4 explicitly covers nonhydroalcoholic fine fragrance, and a body oil falls into the stricter Category 5A. A traditional attar, essentially a distilled natural carried in sandalwood, cannot satisfy a limit of a tenth of one per cent on any ordinary reading of the rules. In Kannauj, where attar has been distilled by the deg and bhapka method for centuries and where Kannauj Perfume holds a geographical indication registered in 2014,17 in the Gulf, where mukhallat and dehn al oud are worn neat, and across the oil traditions of Java and the Malay world, the standard was simply written for a different object and the tradition carried on.
What the rules cost is easiest to see where they are absent.
A small number of independent houses build with materials the mainstream cannot buy. They are not a movement and they do not have a manifesto. What they are is a control group: proof that the missing register is still makeable, and a running demonstration of what the compliant part of the industry is no longer permitted to attempt.
The only house we can document as having openly marketed on non-compliance. It traded from 2017 under the motto Non IFRA-Compliant Fragrance Art
, which Cross later reconsidered for fear it read as indifference to safety, replacing it with Bureaucracy Destroys Art!
He built his own oakmoss substitute out of myrrh and kelp and brought it to a reporter to smell. Neither motto appears on the site today.
Luca Turin's assessment is worth having next to the polemic: solidly built, not especially ambitious, made from a sensible mixture of good naturals and synthetics. He likens it to fragrance played on original instruments, and notes the coincidence of vinyl disappearing at the same moment as oakmoss.
Washington Post Magazine, 21 April 2021, and Luca Turin, 13 April 2024. Source
The house states on its own ingredients page that it uses legally obtained, macerated deer musk grains from Siberia
. Its published note lists include oakmoss absolute, oakmoss resinoid and Russian castoreum. Whole releases are named for the materials: Siberian Musk, Russian Musk, Atlantic Ambergris, Cuir de Russie, Civet de Nuit. Florals are dosed at levels a compliant brief would not survive.
One point the debate consistently gets wrong. Deer musk is not an IFRA matter at all. It appears nowhere in the Standards index. It is governed by the Convention on International Trade in Endangered Species and by national wildlife law, which is a real and serious constraint, and has nothing to do with fragrance allergy. The house ships worldwide, including into Europe.
Areej Le Doré, ingredients page, and Kafkaesque, 31 January 2018. Source
Writes openly about animalics, and claims a provenance no compliance officer could ever verify: At Ensar Oud, all of our deer musk and ambergris is sourced from the late Sultan Qaboos of Oman.
The house describes working with Tonkin, Kashmiri, Mongolian and Tibetan musk and with ancient ambergris, framing the practice as access to a lost history rather than as defiance of a rulebook.
We print that as the house's claim and not as established fact. It is unverifiable by design, which is itself a comment on how this end of the market operates.
Ensar Oud, on animalics and deer musk. Source
The only formal, organised, named dissent on record. The Guild joined the watchdog Cropwatch in calling for a boycott and a moratorium on the adoption of IFRA's 40th Amendment, gathering signatures on a public petition. It was a demand for a pause, not a declaration of non-compliance, and it did not succeed. Nothing comparable has been attempted since.
Anya's Garden, press release and FAQ on the 40th Amendment boycott. Source
Worth naming because unrestricted is not automatically virtuous. Zoologist states plainly that its products are free of natural animal derived musks, replaced with synthetics for ethical reasons, because it does not want to harm animals in order to smell good. An argument against IFRA is not an argument for every material IFRA touched.
Zoologist Perfumes, about page. Source
Two of those houses have never published a position on IFRA at all. They simply build with what they can obtain and let the note list do the arguing. That silence matters, and it points at the part of this system nobody wants to say out loud.
The certificate nobody audits
Take IFRA entirely at its own word, from its own pages. The Certificate of Conformity is a document established by companies creating fragrance mixtures
. IFRA does not create the certificates
, and there are no certifying companies providing certificates on IFRA's behalf. Every supplier is responsible for establishing and providing its own. The Code of Practice is described as mandatory for members, and a member is deemed in breach when it supplies a product violating a Standard, but the page sets out no audit programme, no monitoring, no analytical verification and no routine sanction.27
A sanction does exist on paper. The bylaws require Regular Members to comply, allow the General Assembly to expel a non-conforming member, and give the Board power to request expulsion on a three quarters vote.31 We looked for a published instance of that power being used against anybody, ever, and found none.
So here is the architecture, stated plainly. The seller of the fragrance writes the declaration. Nobody analyses the finished bottle. The association that publishes the rules explicitly disclaims issuing the paperwork. And across Southeast Asia the whole apparatus reduces to a certificate that must be present in a file.
We are not going to assert that independent brands compounding in house routinely exceed the limits, because nobody has produced evidence either way and this guide does not invent evidence. The narrower claim is the one that bites, and it is very hard to argue with: the system contains no mechanism that would ever find out. Conformity is strongest exactly where somebody downstream demands the paperwork, which means at the large houses selling through retailers who ask for it. A maker who compounds in house and sells direct produces a declaration about their own work and files it. That is not a safety system with a gap in it. It is a filing requirement that has been mistaken for a safety system.
Which leaves the regime with the least attractive property a rule can have. It constrains the people who were never the problem, it cannot detect the people it would need to, and the cost of it falls almost entirely on the composition.
Bureaucracy Destroys Art!Manuel Cross, Rogue Perfumery, as reported in the Washington Post
The vintage register did not vanish. It was priced out of the mainstream.
Be precise about the mechanism, because the popular version of this idea is wrong in a way that matters. It is not scale that exempts a perfumer. It is jurisdiction and distribution channel. There is no small batch exemption in European cosmetic law. A hundred bottle run sold in Paris needs the same Responsible Person, the same notification and the same signed safety report as a run of a million. What working outside the mainstream actually buys is the ability to do without the markets that impose compliance in the first place.
These are the four positions from which a pre-2008 dosage is still lawful. Note what each one costs, because none is free, and none is available to a perfumer with ambitions for a department store counter.
The private commission
A bespoke perfume made for one client, not placed on any market. The freest position in the entire craft, and the oldest.
The attar and mukhallat trade
Kannauj, the Gulf, and the oil traditions of Java and the Malay world, working under national cosmetic law that asks for an ingredient list and imposes no allergen declaration.
The domestic independent
A small maker shipping within a market that restricts none of it, as in the United States, where a business under one million dollars in sales is also exempt from federal registration and listing.
The house that argues its own dose
A European maker that commissions its own safety assessment and defends a dosage above an IFRA Standard, where no Annex II or III limit applies. Lawful, and almost never attempted.
State the consequence without sentiment. An entire expressive register has been moved out of the reach of anybody operating at scale, and survives only where the compliance machinery never engages. That is not a romantic fact about artisans. It is a loss of range in the mainstream of an art form, and the people it hurts most are the ones with the resources to have used it well.
It needs three honest qualifications, because an unrestricted palette is a necessary condition and not a sufficient one.
First, some materials are gone commercially rather than legally. Nobody manufactures nitro musks at industrial scale any more, whatever the law says. Thierry Wasser has put this more bluntly than any critic of IFRA: even wanting to reconstruct a vintage Guerlain as it was originally made, the raw materials are not there to buy, and the oils, the oil companies and the growers are long gone or discontinued.
Second, the naturals themselves have changed. Different cultivars, different distillation, different soil, different adulteration. A 2026 rose absolute is not a 1955 rose absolute even where the paperwork says the same thing.
Third, some of what we remember as vintage magic is the bottle and not the formula. A perfume made in 1965 and smelled in 2026 has spent sixty years oxidising, the top has gone and the base has deepened, and what is being admired is partly the ageing. Nostalgia is a fine reason to love something and a poor instrument of measurement.
None of that rescues the regime. Grant all three and the restriction still did what it did, on top of everything else that was already happening, and it is the only one of the four causes that was chosen deliberately by people who could have chosen otherwise.
The deepest cost is the one that never appears in an amendment. Every material on the restricted list is difficult. Oakmoss is bitter and stains, costus smells like a scalp, safrole is medicinal, the rose ketones will swallow a formula whole if you misjudge them by a tenth of a per cent. Learning to place them was most of what an apprenticeship was for. Remove them and you do not merely subtract notes. You remove the possibility of a certain kind of failure, and an art form that cannot fail in interesting ways stops producing interesting successes. Charles Sell told the British Society of Perfumers that it is best to prioritise regulatory compliance over every other aspect, even odour. He was not complaining. He was describing the job.
Give the people the strength to really choose what they want. If you tell them a story, they won't trust you anymore.Andreas Wilhelm, perfumer, Perfume.Sucks
The case for change
Before you blame the rules, check the other four.
An argument for reform is only as good as its diagnosis, and the reformulation debate is full of bad ones. When a perfume you love stops smelling like itself, there are five plausible explanations and only one of them has an acronym.
Supply
A crop fails, a species is listed under an international convention, a producing country changes its export rules, or a single distillery closes. Natural materials arrive from real places with real weather.
Cost
The brief is repriced, the concentrate budget is reduced, and the expensive naturals are the first line item to move. This happens constantly and is never announced.
Ownership
A house is sold, a supply contract moves to a different company, and the formula is rebuilt by a different perfumer working from a different palette of captives.
Creative revision
Somebody simply decided to change it. Femme lost its cumin in 1989 because a perfumer wanted it gone, which is a considerably more interesting fact than a regulation would have been.
Regulation
Real, documented, consequential, and the only one of the five a brand can name out loud without embarrassment. Which is precisely why it absorbs blame far beyond its share.
Regulation is the only suspect in the line-up without a publicist, and it has been quietly confessing to other people's crimes for twenty years. That is a reason to argue against it carefully. It is not a reason to stop.
The remedy that actually fits the risk.
Here is the disproportion, in one paragraph. Fragrance contact allergy affects somewhere near two per cent of adults on a clinically relevant definition, and under one per cent on a strict one. It is uncomfortable, it is persistent, and it is genuinely miserable for the people who have it. It is also not anaphylaxis, it does not kill anybody, it is diagnosable by a patch test that dermatology clinics perform every day, and once diagnosed it is managed entirely by avoidance. Avoidance requires exactly one thing from the industry: that the label say what is in the bottle.
Instead, the material is removed. Not labelled, not dosed down for the sensitive, not confined to products that sit on skin for eight hours. Removed, for everybody, permanently, on the authority of a private association whose science is funded by dues scaled to the sales of the companies being assessed, published behind a subscription, and certified by a piece of paper the seller writes themselves.
Peanut allergy is more common than fragrance allergy and considerably more dangerous. Nobody responded by banning peanuts. They responded by putting it on the packet.
A rare, non-fatal, avoidable condition calls for disclosure and informed choice. It does not call for the permanent abolition of a material from the entire market.
What follows is not an argument for no rules. Prohibitions resting on genotoxicity or reproductive toxicity are a different question and are not disturbed by any of this. It is an argument about the sensitisation restrictions, which are the great majority of the library and the whole of its effect on the art. Five changes would answer them.
Publish the evidence
The RIFM database should be open. A rulebook that a whole industry must live under ought to rest on data any dermatologist, journalist or perfumer can read without a subscription and a price on application.
Fund the science independently
Dues scaled to the sales of the party being assessed is a structure no other field of consumer safety would tolerate. A levy administered at arm's length, or public funding, would cost the industry the same money and remove the objection entirely.
Separate the boards
No individual should sit simultaneously on the body that generates the evidence and the body that writes the rules from it. This is the cheapest reform on the list and the one that would do most for the association's credibility.
Restore fine fragrance as its own case
Aggregate exposure charges a small, deliberate, self-selected use against a daily total dominated by detergent, shampoo and fabric softener. A perfume worn on purpose, by an adult who chose it, is not the same exposure as a laundry rinse and should not be assessed as though it were.
Label, do not remove
Declare the material, at whatever strength the perfumer intends, in plain language on the box. Let the two per cent who need to avoid something avoid it, exactly as they already do with nickel, latex, gluten and peanuts. Everyone else keeps the perfume.
There is a serious objection to the last of those and it should be stated. Labelling puts the burden on the consumer, and a person who does not yet know they are sensitised cannot avoid what they have not been diagnosed for. That is true, and it is the strongest argument the other side has. It is also an argument for better dermatology and clearer labels, not for removing materials from people who are not sensitised and never will be. We accept that trade in every other consumer category on earth. Fragrance is the one place where the answer was to take the thing away.
Readers who disagree are invited to be specific about which of the five they object to. That would be a better conversation than the one the industry has been having, which mostly consists of perfumers complaining privately and an association issuing press releases about consumer confidence.
Constraint can produce invention. It should not be confused with proof that nothing was lost.