Mostly salt, heat and sodium carbonate — and only then a few hundred micrograms of aroma compounds. The citrus-floral top note is real and has been measured: linalool, linalyl acetate, citronellol and geraniol, in a named snus product. But bergamot is named in no primary document for any modern snus, and juniper in none at all.
This is a question where the folklore and the documents part company almost immediately. Search for what gives snus its flavour and you will be told bergamot and juniper, more or less everywhere. I went looking for a primary source that names either for an actual product, and the honest answer is that the citrus half has a sort of ancestor and the juniper half has nothing behind it whatsoever. What is documented turns out to be more interesting: the biggest flavour ingredient in a can of snus, by a factor of roughly 170, is table salt.
Salt is the biggest of these and it has its own page now, with the measured figures by product type, how much of it actually reaches your tongue, and why two people get different amounts out of the same pouch: why some snus tastes of salt.
Salt and soda first — the two biggest flavour ingredients
Swedish Match’s patent application for pouched snus formats sets out what goes into the blend and why. On salt, at paragraph [0054]:
Salt, such as sodium chloride, potassium chloride, magnesium chloride, calcium chloride and any combinations thereof, is added mainly for its effect on taste but it also has a preservative action which contributes to improved shelf life of the product. … Normally, the amount of added salt in the smokeless tobacco composition is within the range of from about 0.5 to about 10% w/w based on dry weight of the tobacco composition.
“Added mainly for its effect on taste.” Not as a preservative with a taste side-effect — taste first, in the company’s own filing. Its consumer FAQ says the same thing more plainly, listing “Cooking salt” as an ingredient that “gives taste and has a preserving effect”, and its own scientists wrote in a peer-reviewed paper that “Sodium chloride adds flavor and prevents microbial activity.”
Put a number on it and the scale becomes clear. In a measured one-gram snus pouch, chloride came in at 35.3 mg — about 58 mg of sodium chloride, roughly 5.8% of the pouch by weight, which lands inside the patent’s stated range. The four named aroma compounds in the same pouch total about 347 micrograms. That arithmetic is mine, not published, but the inputs are two independent sources that agree: salt outweighs every named aroma compound in snus by something like 170 to one.
Nobody writes about salt as flavour, and the manufacturer’s own patent says it is the main one. That is the first correction this topic needs.
Sodium carbonate is the other one. It is usually explained as the thing that raises pH and therefore the hit, which it does. The next patent paragraph, [0055], gives it a second job:
pH adjusters, such as sodium carbonate, are added to bring the pH value to the slightly alkaline side, such as about pH 7.5 to 8.5. Sodium carbonate may also be used to give the products their characteristic aroma profile. Typically, the amount of pH adjuster in the smokeless tobacco composition is less than about 7% w/w, such as within the range of from 3 to 5% w/w, based on dry weight of the tobacco composition.
At 3–5% by dry weight that is another ingredient present in hundreds of times the quantity of any flavour compound, described in the maker’s own filing as contributing the characteristic aroma. If you have noticed that snus has a soapy, mineral, faintly alkaline backbone under whatever is on the label, that is where it comes from. The flavourings sit on top of salt and soda, not on neutral ground.
What has actually been measured
There is exactly one study I can find that quantifies named flavour compounds in a named snus product. It is a British American Tobacco study of a BAT subsidiary’s own product, which is worth knowing before reading it, and the authors declare it: “All authors are employees of British American Tobacco, which is a producer of snus.” The product was “Lucky Strike Original pouched snus (LS Brown) … purchased from Fiedler & Lundgren (Malmo, Sweden)”. Its design was about extraction during use rather than flavour as such, but it had to measure the compounds first.
From its Table 2, per one-gram pouch, with the proportion each gave up over an hour of use:
| Constituent | In the unused pouch | Mean % extracted in 1 h | What it tastes of |
|---|---|---|---|
| Chloride | 35.3 mg | 27.7% | Salt |
| Propylene glycol | 31.1 mg | 29.7% | Humectant, not a flavour |
| Sodium | 24.7 mg | 24.8% | Salt and soda |
| Nicotine | 9.6 mg | 33.3% | — |
| Linalyl acetate | 150.1 µg | 19.2% | Bergamot’s principal ester; sweet, citrus-floral |
| Linalool | 148.8 µg | 30.4% | Floral-citrus, lavender-adjacent |
| Citronellol | 32.5 µg | 28.2% | Rose |
| Geraniol | 15.4 µg | 23.7% | Rose, geranium |
Two of those four are rose markers and two are citrus-floral. Now set that beside what Swedish Match says it adds, unprompted and in an unrelated document: “We add a variety of flavors and aromas to snus, like licorice, lemon oil, and rose oil.” Rose oil is where citronellol and geraniol come from; lemon and bergamot oil both run on linalool and linalyl acetate. The measured compounds and the manufacturer’s named oils corroborate each other, from two unrelated companies — a better story than bergamot, because it is sourced at both ends. It also names the perfumey quality people notice in classic blends and cannot place. It is rose.
Note too that linalyl acetate had the lowest extraction of anything measured, 19.2%, because it is the least water-soluble. Flavour does not leave a pouch at the rate the nicotine does, which is part of why a portion tastes strongest early and flattens while still working.
The bergamot question — and the juniper one
Bergamot does appear in a primary source. Once. In a paper by Swedish Match’s own scientists, in a passage about Jacob Ljunglöf’s nineteenth-century recipe and his contemporaries:
Most manufacturers used flavorings, such as, salmiac, pepper, and lemon or bergamot oil [I. Junhem, personal communication].
Three things disqualify that as a product fact, all visible in the same paragraph. It is explicitly historical — the surrounding sentences describe “recipes and production techniques used by other manufacturers at the time”, and the next one begins “During the 20th century production of snus became more centralized.” It names no product. And the citation resolves, through the acknowledgements, to the director of the Museum of Tobacco and Matches in Stockholm, thanked “for providing access to the Museum’s archives”. The entire primary-source basis for bergamot in snus is a museum director’s account of 1800s recipes. The authors’ own declaration: “All authors are or have been employees of Swedish Match AB.”
Today’s picture is different. Bergamot appears nowhere on Swedish Match’s current site — not the FAQ, not the production pages, not the sitemap — where the named oils are liquorice, lemon and rose. It appears in neither of the company’s relevant patent families, and in no regulator document I read.
There is also a chemical argument against reading the measured compounds as neat bergamot oil, and it cuts against what I expected to find. The European Medicines Agency’s assessment of bergamot oil gives linalool at 6.0–15.0% and “Linalyle acetate” (the spelling is theirs) at 23.0–35.0%, and notes that “Oxygenated compounds, namely linalool and linalyl acetate, mark the flavour notes of BEO”. That is a linalool-to-ester ratio of roughly 0.2 to 0.65. The measured snus pouch is 148.8 against 150.1 — essentially one to one, outside bergamot’s range and closer to lavender, petitgrain, or a compounded blend built to a specification rather than a single pressed oil.
So the compounds are consistent with a bergamot-type note, the ratio argues against bergamot oil itself, and no document names it for a modern product. A pleasing footnote: linalyl acetate’s trade synonyms in the WHO’s flavouring database include “BERGAMOL”. The compound carries the name even where the oil does not. For scale, linalool, linalyl acetate and citronellol share a WHO group acceptable daily intake of 0–0.5 mg per kg of body weight, against roughly 88 micrograms actually extracted from that pouch — a low single-digit percentage of it at most, and I am not computing a precise figure because the ADI is expressed as citral and the conversion is not in the literature.
Juniper is simpler: there is nothing. I searched for juniper, and for the Swedish enbär, across both Swedish Match patent families, every page on the company’s own site including its sitemap, LIVSFS 2024:2, the Swedish Food Agency’s guidance, the Swedish nicotine statutes, the EU Tobacco Products Directive, and all three of the analytical papers above. Zero hits. Everywhere.
That is the strongest kind of negative available: not “I could not confirm it” but “I looked in every document that would carry it and the word is absent.” Juniper is a real flavouring and a plausible one for a Nordic product, and the EMA publishes a perfectly good composition for juniper berry oil — α-pinene at 20 to 50%, β-myrcene at 1 to 35% — though those are pharmacopoeia specification limits, not measured ranges. None of it connects to snus. If a maker uses juniper, it has not said so, and nobody has measured it.
The smoke is not smoke, and the heat step is a flavour step
The leathery, smoky, slightly tarry note in traditional blends has a documented history, and it is the best unexpected thing in this research. It used to be real. It is not any more.
Traditionally snus was made from a selection of tobacco varieties typically including a large proportion of fire-cured tobacco. This resulted in relatively high levels of PAH:s in the finished products. During the 1990s use of fire-cured tobacco was phased out.
Fire-curing hangs leaf in smoke, which is where the flavour came from and also the polycyclic aromatic hydrocarbons; the same paper records benzo(a)pyrene falling from about 20–25 ng/g to under 2–3 ng/g on a dry-weight basis as a result. And then: “Fire-cured tobacco adds a distinct flavor, but the traditional flavor characteristics were preserved using food-approved flavorings”, and “In part, flavoring is used to compensate for the fact that fire-cured tobacco is no longer used in snus”.
The smoky note in modern snus is an added flavouring imitating a curing method the industry abandoned for carcinogen reasons. It is a deliberate reconstruction of a taste whose original source was the problem. That belongs alongside the nitrosamine story as one of the two big chemistry changes of the 1990s.
The Swedish Food Agency closes the loop on it, and adds a deadline. Smoke flavourings are expressly permitted in snus, and those that did not get renewed EU authorisation “får fortsätta användas i tillverkningen av snus och tuggtobak så länge som de är godkända att användas i något livsmedel, det vill säga till 1 juli 2029.” Many of the smoke flavourings now standing in for fire-curing have until 1 July 2029. I am attributing that date to the Agency’s guidance, because the underlying EU implementing regulation would not load on any route I tried.
The heat step is the other process contribution. Swedish snus is heat-treated rather than fermented, which is what distinguishes it from American moist snuff, and the purpose is usually given as microbiological. The company’s own description puts taste first:
The mixture is then subjected to a computer-controlled heat-treatment process, the purpose of which is to improve the taste and to reduce microbial activity to obtain a proper shelf life.
The patent agrees from the process side: the blend is held hot “to bring texture, taste and color to the snus blend.”
What I cannot tell you is which chemistry. Heating a wet, alkaline, sugar-and-amino-acid mixture for hours is the textbook condition for Maillard reactions and pyrazine formation, and that is the obvious explanation. It is also unsourced: PubMed searches for snus with Maillard, snus with pyrazines, and smokeless tobacco with pyrazines all return zero, and the independent flavour panel above did assay 2,3,5-trimethylpyrazine and report it in no product at all. The heat step changes the taste, the maker says so, and nobody has published what it forms. I am not going to fill that in with plausible chemistry.
The one non-industry measurement of flavour compounds across oral products is a US National Cancer Institute-funded study that compared tobacco-free pouches with eleven oral tobacco products from seven brands, General among them. Its panel of 33 flavouring chemicals included both bergamot markers by name: “limonene, linalool, maltol, menthol, menthone, menthyl acetate, methyl salicylate” and the rest.
Linalool appears exactly once in that entire paper — in the list of what they looked for — and in no result. Limonene appears only as a finding in one pouch flavour capsule, never for any oral tobacco product. An independent lab assayed for the two compounds that mark a bergamot-type note, across eleven oral tobacco products including General, and reported neither.
What it did find in oral tobacco products was cooling and wintergreen chemistry: “menthol was identified most frequently in OTPs (45.5%)”, with the highest single concentration a wintergreen moist snuff at 26.8 mg of methyl salicylate per pouch. And “Benzyl alcohol, which provides a fruity taste” turned up in General’s mint variety.
Two caveats. This was a US-market sample, and US-market General is not necessarily the Swedish blend; and its per-product tables sit in supplementary files I could not retrieve, so I cannot give you General’s own numbers. The direction is clear enough, though: the modern mass market, oral tobacco included, runs on mint and cooling agents rather than perfumery, and cooling does something specific to how strong a product feels. The whole flavour landscape is a separate piece.
Who has to tell you what, and the asymmetry inside one regulation
Here is the part that decides how much of this you can ever check for yourself. Both snus and tobacco-free pouches are reported to the state in full, with quantities. Neither is disclosed to you in full. But only one of them has to print anything on the tin.
On reporting, the Tobacco Products Directive requires manufacturers to submit, by brand name and type, “a list of all ingredients, and quantities thereof, used in the manufacture of the tobacco products, in descending order of the weight of each ingredient”. Sweden’s tobacco-free nicotine regime requires the equivalent for pouches — every ingredient “uppdelad på varumärke och sort, med uppgift om kvantiteter”, broken down by brand and variant, with quantities. Both regimes also carry the same escape hatch. The Directive says member states “shall take the need to protect trade secrets duly into account when making that information publicly available”; the Swedish ordinance publishes what “inte utgör företagshemligheter”. A flavour formulation is the textbook trade secret, which is where it disappears. Swedish Match says as much itself: “the exact recipes are business secrets, of course.”
On labelling, the asymmetry is stark, and the cleanest version of it is inside a single regulation. LIVSFS 2024:2, the Swedish Food Agency’s rules for snus, requires four things on a can sold to a consumer: net quantity in grams, date of manufacture, a storage instruction where it matters for shelf life, and the maker’s name and address. No ingredients, no flavour. Four sections later, the same regulation requires a nicotine-free snus-like product to carry “en förteckning över alla ingredienser som ingår i produkten i fallande viktordning” — every ingredient in descending weight order. And a nicotine pouch must carry an ingredient declaration too, under Lag (2022:1257). The Agency spells out the three-way split itself: a nicotine-containing snus-like product needs neither a product designation nor an ingredient list under its rules, only the four items above.
So, in the Swedish cabinet: the nicotine-free lookalike is the most disclosed product, the nicotine pouch next, and the tobacco snus — the one with the longest history and the agency that has regulated it the longest — tells you the least. One document, three tiers, no argument available.
One footnote on provenance, because it bears on how much weight the company’s own pages carry. Swedish Match cites a “Swedish Match Ingredients Policy” in both its FAQ and its peer-reviewed literature; the URL for it now returns a page of financial reports. Cited twice, published nowhere. Its FAQ also attributes the propylene glycol limit to LIVSFS 2012:6, repealed on 1 April 2024 — the number survived into the current annex at 40 g/kg, but the citation is stale.
Why snus escaped the EU flavour ban, and what governs it instead
Flavour is the first thing regulators reach for, and snus has a curious position. The Directive’s Article 7(1) says member states “shall prohibit the placing on the market of tobacco products with a characterising flavour” — but Article 7(12) exempts “Tobacco products other than cigarettes and roll-your-own tobacco”. Snus is out. Its enhanced additive reporting in Article 6 is likewise limited to cigarettes and roll-your-own, so the one place the EU demands deep additive studies, snus sits outside that too.
What constrains snus flavour instead is food law, and only in Sweden, because Article 17 bans oral tobacco everywhere else in the Union — the exemption Sweden negotiated on accession is why the product still exists. The operative rule is LIVSFS 2024:2, 5 §:
Endast sådana aromer, inklusive rökaromer, och livsmedel med aromgivande egenskaper som får användas i livsmedel, får ingå i snus och tuggtobak för att ge eller förändra lukt eller smak.
Only flavourings — smoke flavourings included — that may be used in food may be used in snus to give or alter smell or taste. It is a positive-list regime borrowed wholesale from food, which is why a snus flavouring is a food flavouring and why the ingredient annex reads like an additives schedule.
And a nuance worth getting right, since the claim circulates in a sloppier form. Snus is not legally food in Sweden. The Food Agency states plainly that snus and snus-like products “är inte livsmedel enligt definitionen av livsmedel i artikel 2 i förordning (EG) nr 178/2002”, and that EU food legislation therefore cannot be applied to them. What the Swedish Food Act does instead is equate snus with food — “jämställts med livsmedel” — as what the Agency itself calls a “lagteknisk lösning”, a legal-technical device, so that selected parts of food law can reach the product. If you see “snus is classified as food in Sweden”, that is an over-simplification of a real statutory mechanism rather than an invention. The accurate version is that Swedish law applies parts of food law to snus by treating it as if it were food.
Where the flavour comes from, in one table
| Ingredient | What it contributes | Where it is named | Quantity published? |
|---|---|---|---|
| Sodium chloride | Taste, “mainly”; water activity | Patent [0054]; maker’s FAQ; maker’s own paper | Yes — 0.5–10% w/w; 35.3 mg chloride per pouch measured |
| Sodium carbonate | pH 7.5–8.5 and “characteristic aroma profile” | Patent [0055] | Yes — 3–5% w/w, under 7% |
| Heat treatment | “To improve the taste”; “texture, taste and color” | Maker’s own paper; patent [0111] | No — chemistry unpublished |
| Linalool, linalyl acetate | Citrus-floral top note | Measured in a named snus; EMA bergamot assessment | Yes — 148.8 and 150.1 µg per pouch |
| Citronellol, geraniol | Rose | Measured in a named snus; corroborates “rose oil” | Yes — 32.5 and 15.4 µg per pouch |
| Liquorice, lemon oil, rose oil | Sweet-anisic, citrus, floral | Swedish Match FAQ | No |
| Smoke flavourings | Stands in for withdrawn fire-cured tobacco | LIVSFS 2024:2, 5 §; rationale in maker’s own paper | No — phase-out dated 1 July 2029 |
| Propylene glycol | Humectant, not a flavour, but mass-dominant | LIVSFS 2024:2 annex (E 1520, max 40 g/kg) | Yes — 31.1 mg per pouch measured |
| Bergamot oil | Earl-Grey citrus | One 19th-century mention, via personal communication, no product named | No |
| Juniper | — | Nowhere | No |
| Pyrazines / Maillard products | — | Nowhere; assayed by one lab, found in nothing | No |
Frequently asked questions
Does General snus actually contain bergamot? No primary document says so. Swedish Match’s current list of named flavourings is liquorice, lemon oil and rose oil. A measured snus pouch does contain linalool and linalyl acetate, which are what bergamot’s flavour runs on, but the ratio between them does not match bergamot oil, and lemon oil carries the same compounds. The honest answer is a bergamot-type citrus note from compounds that have been measured, not bergamot oil that anyone has documented.
Why does snus taste salty? Because it is. Roughly 5–6% of the pouch weight in the one product anyone has published figures for, and the patent describes salt as added mainly for taste. Salt also tastes stronger as a pouch dries out, which is part of why moisture changes so much.
Why does the flavour fade before the pouch is finished? Different compounds leave at different rates. Linalyl acetate gave up 19.2% of itself in an hour against nicotine’s 33.3%, because it is far less water-soluble. The aroma is loaded in micrograms and front-loads; the nicotine keeps going. So the taste flattening is not a sign the pouch is spent.
Is the smoky flavour in some snus real smoke? Not since the 1990s. Fire-cured tobacco was phased out for carcinogen reasons and the flavour is now built with added smoke flavourings — which themselves have a 1 July 2029 deadline in Sweden for those that lost EU authorisation.
Can I find out what is in my can? Not from the can. Swedish rules require net weight, manufacturing date, a storage note and the maker’s address — no ingredient list. The full recipe goes to the state and is withheld from publication as a trade secret. Ironically, a nicotine-free snus-like product in the same regulation must list everything. Reading what a can does tell you is a separate skill.
Are flavoured snus varieties regulated differently from flavoured cigarettes? Yes, and much more loosely. The EU characterising-flavour ban exempts everything other than cigarettes and roll-your-own, so snus flavour is governed by food-additive rules rather than tobacco flavour rules.
What about flavour in tobacco-free pouches? Different legal regime, different chemistry, and much better documented — including the cooling agents that make a pouch feel stronger than it is. We have also looked at whether pouches are gluten-free, which turns on the same disclosure gap from the other direction.
Is any of this in GothiaTek? No. The voluntary standard is a ceiling on undesirable substances — the patent describes it as implying “hygienic handling of all ingredients and pasteurization of the loaded materials” — not an ingredient-origin or flavour standard. It does not help with this question at all.
Sources
- Rutqvist LE, Curvall M, Hassler T, Ringberger T, Wahlberg I. “Swedish snus and the GothiaTek® standard.” Harm Reduction Journal 2011;8:11, PMC3119032. Historical flavourings, the fire-curing phase-out and the heat-treatment purpose. Declared conflict: “All authors are or have been employees of Swedish Match AB.”
- Digard H, Gale N, Errington G, Peters N, McAdam K. “Multi-analyte approach for determining the extraction of tobacco constituents from pouched snus by consumers during use.” Chemistry Central Journal 2013;7:55, PMC3618139. Table 2 figures. Declared conflict: “All authors are employees of British American Tobacco, which is a producer of snus.”
- Page MK, Leigh NJ, et al. “Comparison of Nicotine and Selected Flavoring Contents Between Tobacco-Free and Tobacco-Containing Oral Pouches.” Nicotine & Tobacco Research 2026, PMC12824957. NCI-funded; the senior author’s consultancies are declared in the paper.
- Swedish Match, European patent application EP3087852A1, paragraphs [0042], [0054]–[0057] and [0111].
- European Medicines Agency, HMPC assessment report on Citrus bergamia Risso et Poiteau, aetheroleum, EMA/HMPC/56155/2011, and the corresponding report on Juniperus communis L., pseudo-fructus.
- WHO/FAO JECFA, linalyl acetate (JECFA 359, CAS 115-95-7), group ADI 0–0.5 mg/kg bw.
- Livsmedelsverket, LIVSFS 2024:2, 5 §, 12 §, 14 § and annex; and Kontrollwiki article 887 on snus, snus-like products and chewing tobacco.
- Livsmedelslag (2006:804), 3 §; Lag (2022:1257) 7 §; Förordning (2022:1263) 4 § and 6 §.
- Directive 2014/40/EU, Articles 5, 6, 7 and 17 (European Commission’s own copy).
- Swedish Match corporate FAQ, read 30 October 2026.
