Musk in Perfume After the Nitro-Musks

Musk on a modern label means a synthetic. Natural deer musk has been under international trade control since 1979 and was never affordable at scale. The nitro musks that replaced it were themselves restricted or prohibited, and today’s musks are polycyclic and macrocyclic materials with a very different smell.

Musk is the third most frequently mentioned material in our own catalogue’s description copy, appearing in 316 listings, behind only bergamot and rose at 334 each. It is also the word on a perfume box that has drifted furthest from what it originally described. Nothing sold today smells much like the thing the word was coined for, and that is the result of three separate waves of substitution over roughly 140 years. This is what happened, and what “musk” is actually doing in your bottle now.

What was natural musk?

A glandular secretion from the male musk deer, dried into a grain and used as both a scent and a fixative. A 1999 report by Volker Homes for TRAFFIC Europe, On the Scent: Conserving Musk Deer, is the most useful single document on the trade, and its numbers explain why the material was always going to be replaced.

• The report states that “musk is used not only as a fragrance but also as a fixative for other fragrance” — that second job is the important one.
• It calculates that approximately 160 deer in total are killed per kilogram of musk, reasoning from three to five animals taken per male with a usable gland and forty such males required per kilogram.
• It records a market value reaching US$45,000 per kilogram at the end of the 1970s, and European retail prices of roughly US$30 to US$50 per gram in the 1990s.
• It reports that France imported approximately 97 kg of unprocessed musk from 1980 to 1995 — a figure that gives a sense of how small the legitimate European trade already was.
• It states that use of natural musk in the European perfume industry declined during the 1990s.

On trade controls, the same report records that Himalayan-region populations of the Siberian musk deer, Moschus moschiferus — specifically those occurring in Afghanistan, Bhutan, India, Myanmar, Nepal and Pakistan — have been listed in CITES Appendix I since 1979, with all other musk deer species in Appendix II since the same year. CITES appendices are amended over time, so anyone needing current status should check with CITES directly rather than relying on a 1999 report.

A separate figure from the trade press puts the historical scale in perspective. A 1992 article in Perfumer & Flavorist by Danute Pajaujis Anonis reports that statistics from the 1930s indicated about 2,400 kg of musk was needed annually to meet demand in “the Orient and other parts of the world.”

Put the price and the yield together and the conclusion is unavoidable: natural musk was never a mass-market material. Even before any trade restriction, a fragrance selling at scale could not be built on something costing tens of thousands of dollars a kilogram.

The first synthetic musks, and what happened to them

The nitro musks arrived in the 1880s and dominated for a century. The IARC monograph on musk ambrette and musk xylene, volume 65, published in 1996, gives the origin: Albert Baur is credited with discovering and commercialising them, and in his 1889 German patent “Baur described a process whereby toluene was butylated with butyl halide in the presence of aluminium chloride,” with the resulting product nitrated to produce musk-scented compounds. By 1892 Baur had identified his original creation as trinitro-butyltoluene and had also developed musk ambrette by nitrating butylated meta-cresol methyl ether.

The same monograph gives the scale, and it is worth quoting because dated production figures for aroma materials are rare:

Roughly 2,450 tonnes down to about 1,000 in a few years, a fall of around 60 percent on IARC’s own two figures, is not ordinary market drift. Two regulatory stories sit behind it.

Musk ambrette

IARC records that following photosensitivity reports in 1979 and 1980, IFRA issued guidelines in 1981 limiting musk ambrette to 4 percent in fragrance compounds; that by 1983 IFRA recommended discontinuing its use in cosmetics and toiletries in contact with skin; and that in the United States musk ambrette was removed from the GRAS list in 1984. Musk ambrette now appears in the IFRA Standards index as a Prohibition.

Musk xylene

This one went through European chemicals law rather than cosmetics law, and the route matters. ECHA identified musk xylene — CAS 81-15-2, EC 201-329-4 — as a substance of very high concern on the ground that it is very persistent and very bioaccumulative under Article 57(e) of REACH, adding it to the Candidate List by decision ED/67/2008 of 28 October 2008. Commission Regulation (EU) No 143/2011 of 17 February 2011 then moved it to Annex XIV, the authorisation list, as entry 1, “5-tert-butyl-2,4,6-trinitro-m-xylene (Musk xylene),” with a latest application date of 21 February 2013 and a sunset date of 21 August 2014.

Note the ground: persistence and bioaccumulation in the environment, not skin. Premium Beauty News, reporting on 24 February 2011, noted that IFRA had already banned the material and that it “once ranked among the most widely used synthetic musks in the perfumery industry before declining in the early 1990s.” ECHA’s background document on musk xylene, dated 1 June 2009, records the assumption made in a 2008 consultants’ report that 80 percent of the overall tonnage went into detergents, cleaning products and fabric softeners and 20 percent into toiletries, colognes and shampoos — so perfume was never the main consumer of it.

Musk xylene, musk ambrette, musk tibetene and musk moskene all appear in the current IFRA Standards index as Prohibitions. Musk ketone appears with a Specification. Anyone saying “the nitro musks were banned” is roughly right about the outcome and usually wrong about the instrument: some of it was IFRA, some was EU chemicals law, some was US food regulation, and the dates are spread across three decades.

What replaced them?

Two chemical families, with quite different characters.

Polycyclic musks. The dominant workhorses of the late twentieth century. The 1992 Perfumer & Flavorist article lists Galaxolide as an isochroman musk from IFF that achieved commercial success, and Tonalide as among the first bicyclic musks to gain wide use, from PFW. That article gives no dates for either, and we are not supplying any, because we could not source them.

Macrocyclic musks. Large-ring molecules closer in structure to the natural material. The same article records that Waldbaum had recognised muscone as the principal odoriferous component of Tonquin musk as early as 1906; that “Ruzicka and coworkers established the synthesis of Exaltone (cyclopentadecanone) and Exaltolide”; and that “Ambrettolide, a 17-membered ring macrocyclic lactone with a musk-type odor, was discovered and synthesized by Kerschbaum.” No dates are given for the Ruzicka or Kerschbaum work in that source, so none appear here.

The article’s own summary of why the substitution happened is refreshingly unromantic: the new synthetic musks, it says, were being used more and more boldly in fragrances because of their greater stability and lower prices. That was written in 1992, and it describes the mechanism for every substitution in this article.

Are today’s musks under pressure too?

Almost certainly, and we are going to say so without naming a live proceeding. Hazard classifications and use limits for individual fragrance materials move continuously through the EU’s CLP process and through IFRA’s amendment cycle, and a material’s status can change between one amendment and the next. We are a fragrance retailer rather than a toxicologist, and reporting a part-finished classification dossier accurately would mean publishing a snapshot that is out of date before you read it. For the current position on a specific musk, read the CLP registry and IFRA’s own Standards Library rather than a fragrance article or a forum.

The general pattern, though, is worth naming plainly: every musk in commercial use has eventually been replaced by a cheaper, more stable or less constrained one, and there is no reason to expect the current generation to be the last. Why perfumes get reformulated explains how that process reaches your bottle, usually without announcement.

What do modern musks actually smell like?

Our characterisation, from handling the finished fragrances rather than the raw materials in isolation.

• Clean rather than animalic. This is the single biggest shift. Modern musks read as laundry, warm skin, soap and cashmere. The word “musk” in nineteenth-century perfumery meant something considerably more animal.
• Structural rather than decorative. A musk’s main job is usually to extend and blur everything else — the fixative role the TRAFFIC report identified in the natural material. Most of what a musk does in a fragrance, you do not notice as musk. The dry woody-amber materials described in Iso E Super and the woody masculine do a related structural job.
• Enormously variable between people. Musks are the material class with the best-documented differences in individual perception. We have set out that evidence, including the specific receptor variants and the measured differences in perceived intensity, in our guide to skin chemistry. If a musk-led fragrance smells like nothing to you and wonderful to a friend, that is expected rather than odd.
• The “skin scent” effect. A quiet musk base is one of the main ways a fragrance is made to read as an enhanced version of the wearer rather than as an applied product. Ambroxan-type materials do a related job, described in the ambroxan story.

How should you shop for a musk?

Four practical points.

1. Ignore the word on the box. “Musk” on a listing tells you nothing about which musk, and it never will. Our catalogue has no ingredient field; only about 1.5 percent of listings publish a full note pyramid at all. The five things to read on a listing explains what a product page can and cannot tell you.
2. Test on skin, and give it three hours. A musk’s contribution is mostly in the second half of the wear. Judging a musk on a paper strip in a shop is close to pointless. The testing guide sets out a method.
3. Ask someone else. Because individual sensitivity to musks varies more than for almost any other material class, your own reading is less reliable here than usual.
4. Do not chase “real musk.” It is not a thing you can buy in a modern fragrance, it has been trade-controlled since 1979, and anyone offering it should be treated with suspicion. How to tell a fake from the real thing covers claims that should stop you.

Myths worth correcting

“Expensive perfumes still use real deer musk.” Natural musk has been under CITES control since 1979 and was priced at tens of thousands of dollars a kilogram before that. The TRAFFIC report records European industry use declining through the 1990s. Nothing on a mainstream shelf contains it.

“White musk is a specific material.” It is a marketing term for the clean, soapy, laundered register that the polycyclic and macrocyclic musks produce. It does not identify a molecule.

“The nitro musks were banned for skin safety.” Some of that story is about skin and some is not. Musk ambrette’s restriction followed photosensitivity reports and ran through IFRA guidance in 1981 and 1983 and the US GRAS list in 1984. Musk xylene went a completely different route: it was identified as very persistent and very bioaccumulative in the environment and moved onto the REACH authorisation list.

“Synthetic musks are a modern shortcut.” Nitro musks date from Baur’s work in the 1880s and were standard base materials across the classical era of perfumery. The great fragrances of the early twentieth century were built with them. What naturals and synthetics each actually do makes the general case.

“I cannot smell musk, so my nose is broken.” Reduced sensitivity to specific musk molecules is common and genetically based, and it is one of the best-documented facts in olfaction. It is a variation, not a fault.

“Macrocyclic musks are natural because they resemble the natural molecule.” They are synthesised. Structural similarity to a natural compound is not naturalness, and neither category is safer or less safe than the other — that is not a claim we make or are qualified to make.

The honest limits of this article

This is one of the better-sourced pieces in the series, and it still has real soft spots.

The dates are uneven. IARC volume 65 gives Baur’s 1889 patent and his 1892 identification of the compound. A great many other accounts give 1888 as the year the first nitro musk was discovered. We have used IARC’s dated account here because it is the source we read and because it is the figure tied to a dated patent; our naturals-and-synthetics article has now been corrected to match. A reader should still know that both figures circulate. The 1906 muscone attribution to Waldbaum, and the Ruzicka and Kerschbaum work, come from a 1992 Perfumer & Flavorist article that gives no dates for the latter two; we have not supplied any.

Galaxolide and Tonalide have no dates or inventors here beyond the companies named in that 1992 article, because we could not source them. Note also that the only copy of that article available to us is a scan whose character recognition is visibly degraded. Where a sentence mattered we have quoted it only where the extracted text was clean, and otherwise reported it rather than quoted it.

The musk xylene dates required care. The Annex XIV entry in the original text of Commission Regulation (EU) No 143/2011 and the entry in the corrigendum that replaced that Annex do not read identically; the dates given above — latest application 21 February 2013, sunset 21 August 2014 — are those printed in the corrigendum, which is the operative version. We flag this because it is exactly the sort of detail that gets repeated wrongly.

The TRAFFIC figures are from 1999 and are dated in the text as such: the US$45,000 per kilogram is an end-of-1970s market value, the US$30 to US$50 per gram is 1990s European retail, and the 160-deer-per-kilogram figure is that report’s own calculation from its own assumptions, not a measurement. The CITES listing detail is as that report states it and appendices change.

We deliberately report no live hazard-classification proceeding. An earlier draft reported a 2025 proposal by a national agency to classify one current musk under the EU CLP Regulation. It was cut at fact-check, because the proceeding had already moved several stages past the point the draft described and its current position could not be confirmed from a primary document. We make no health or safety claim about any musk here, and we do not suggest that any synthetic is safer or less safe than any natural. Where a regulator has stated a ground for a completed action, which for musk ambrette was photosensitivity and for musk xylene was very high persistence and bioaccumulation, we have reported the regulator’s stated ground and stopped.

No fragrance in this article is claimed to contain any specific musk. Formulas are trade secrets. No perfumer is named anywhere. Every olfactory description is our characterisation, from experience of finished fragrances. And the catalogue figure — musk mentioned in 316 listings — counts words in sales copy, not molecules in bottles.

What our own catalogue shows

The one musk figure our data holds is above: musk is the third most-mentioned material in our description copy. Everything else that was asked for here is unavailable, and the unavailability is the point.

We cannot tell you how many of those listings say “white musk” specifically, which is the phrase most likely to be doing the work on a modern label, because the counted vocabulary records the material and not the qualifier. Nor can we count listings that mention musk and no other base material — the nearest thing to a count of fragrances sold as musks rather than fragrances with musk in them — because co-occurrence across that vocabulary was never computed.

The deeper limit is the one this article keeps returning to. Only 61 of our roughly 3,950 listings publish a full top-heart-base pyramid, and 2,814 carry no note text at all. “Musk” on a label is a category word covering several chemical families with different smells, different regulatory histories and different persistence, and no listing in our catalogue tells you which family is in the bottle. A count of the word would be a count of the word.

Related reading

• Naturals against synthetics — why that division misleads buyers, and what either side is really for
• How ambroxan took over the modern masculine — the molecule behind the skin-scent effect in so many recent launches
• Our reformulation guide — IFRA, the EU, and what changed inside the bottle
• Our Swiss Arabian guide — the Sharjah house, what the group owns, and why it costs more

Common questions

Is there real musk in perfume?

Not in anything on a mainstream shelf. Natural deer musk has been under CITES trade control since 1979, and a 1999 TRAFFIC Europe report put its market value at around US$45,000 per kilogram at the end of the 1970s, with European retail prices of roughly US$30 to US$50 per gram in the 1990s. The same report says European perfume industry use of natural musk declined through that decade.

What does musk mean on a perfume label?

A synthetic, almost certainly from the polycyclic or macrocyclic families. The word identifies a register — clean, soapy, skin-like, laundered — rather than a specific molecule, and no listing will tell you which material was used. Musk is the third most mentioned material in our own catalogue’s description copy, in 316 listings, which counts words in sales copy rather than molecules in bottles.

Why were nitro musks banned?

Different materials for different reasons through different instruments. Musk ambrette was limited by IFRA to 4 percent in 1981 after photosensitivity reports, recommended for discontinuation in skin-contact cosmetics by 1983, and removed from the US GRAS list in 1984. Musk xylene went through European chemicals law instead: ECHA identified it as very persistent and very bioaccumulative, and it was moved onto the REACH authorisation list in 2011.

Who invented synthetic musk?

Albert Baur is credited with discovering and commercialising the nitro musks. The IARC monograph on musk ambrette and musk xylene, volume 65, published in 1996, describes his 1889 German patent for a process butylating toluene with a butyl halide in the presence of aluminium chloride and then nitrating the product, and says that by 1892 he had identified his original creation as trinitro-butyltoluene. Note that 1888 is also widely cited as the discovery year.

What is white musk?

A marketing term, not a material. It describes the clean, soapy, laundered register produced by modern polycyclic and macrocyclic musks, in contrast to the more animal character of what musk meant in nineteenth-century perfumery. There is no molecule called white musk, and two fragrances both described that way may share nothing in common beyond the effect.

What replaced the nitro musks?

Two families. Polycyclic musks, including Galaxolide, described in a 1992 Perfumer & Flavorist article as an isochroman musk from IFF, and Tonalide, described there as among the first bicyclic musks in wide use, from PFW. And macrocyclic musks, large-ring molecules closer in structure to the natural material, including muscone, Exaltone and Exaltolide, and ambrettolide. That article’s own explanation for the switch was greater stability and lower prices.

Why can some people not smell musk?

Because sensitivity to specific musk molecules varies between people on a genetic basis, and musks are the material class where this is best documented. It is a variation rather than a fault. If a musk-led fragrance smells like almost nothing to you and wonderful to someone else, that is the expected outcome. It also means your own reading of a musk is less reliable than usual — ask someone.

How many deer were killed for a kilogram of musk?

A 1999 TRAFFIC Europe report calculated approximately 160 deer in total per kilogram, reasoning from three to five animals taken for every male with a usable gland and forty such males needed per kilogram. That is the report’s own calculation from its own stated assumptions rather than a measurement. The same report noted that France imported roughly 97 kg of unprocessed musk from 1980 to 1995.