Ethyl Carbamate: Enforced for Forty Years, Law Since July
Canadian importers of rhum agricole, the French AOC- and IG-bearing cane-juice rums of Martinique and Guadeloupe, have told me more than once that bringing rhum into Ontario is close to a coin flip: roughly half the time, a shipment fails the LCBO's laboratory and never reaches a shelf. That is their estimate, but it’s enough to dissuade importers from doing much rhum agricole business in Ontario. The molecule at fault is ethyl carbamate, a process contaminant that can form during fermentation, distillation and storage, and one the International Agency for Research on Cancer classifies as probably carcinogenic to humans. Historically, the threshold number was 150 micrograms per litre.
On 24 July 2026, that number changed its legal character without changing its value. If you make, import or formulate with alcohol, the change is worth understanding.
A Change to the List of Contaminants and Other Adulterating Substances
Health Canada moved the ethyl carbamate maximum levels onto the List of Contaminants and Other Adulterating Substances in Foods (the “List”), a document incorporated by reference into Division 15 of the Food and Drug Regulations (FDR). Health Canada announced the change in Notice of Modification MCON-2605 (the “NOM”) and the modification took effect on 24 July 2026. The operative entry is item E.1 of Part 2, which sets out the maximum levels of ethyl carbamate for the following items:
(a) From maple sap; from rice; baijiu: 200 ppb
(b) Distilled spirits, except those in (c): 150 ppb
(c) Distilled spirits from fruit; fortified wines aged 20 years or greater: 400 ppb
(d) Fortified wines except those in (c); wines aged 20 years or greater; wines with greater than 60 g/L residual sugar: 100 ppb
(e) Wines, except those in (d): 30 ppb
In a sense, the numbers were not law until July, but - as is common in Canadian food law - these very numbers sat on an administrative list. A result over the maximum gave the CFIA grounds to act, but it did not by itself make the product illegal. The regulator still had to assess the particular product and make the choice as to how to proceed. Health Canada's position is that, while the consolidation proceeds, "both sets of Maximum Levels are enforceable by the Canadian Food Inspection Agency."
FDR’s Subsection B.15.001(2) provides that a food on the List is adulterated where the substance exceeds the maximum level. CFIA doesn’t have to perform an assessment of the particular product, and it doesn’t have to build a case to demonstrate that the amount present is harmful. Over the listed number is adulterated, and adulterated engages the Food and Drugs Act prohibition on sale.
Subsection (3) does the opposite: where the substance sits at or below the maximum level, the food is, in respect of that substance, expressly exempt from paragraph 4(1)(a) of the Act, the poisonous-or-harmful-substance prohibition. Before July, a compliant producer had to hit a number and then make a good argument. It now has an exemption written into the Regulations. So: some added certainty for producers and importers, which is helpful.
Of note is that the category lines were redrawn on the way. Icewine is the leading example: VQA Ontario requires not less than 100 g/L residual sugar in ice wine, and the new paragraph (d) captures wines above 60 g/L, so the whole category moved from 30 (table wines - ice wine isn’t fortified, per se) to 100 ppb by operation of its own standard. The old category also read "fruit brandies and liqueurs" at 400 ppb, and industry comments show many understood that as covering every liqueur. Health Canada says it never did, and 400 ppb applied only to liqueurs made from fruit, and a nut, coffee or cream liqueur was always in the general 150 band. The new wording spells out their position (news to most in industry!) and because Health Canada’s position is that nothing changed, then there is no need for an adjustment period. So, if a supplier certified amaretto to 400 ppb, it was certified to a number that never applied, and any bottle over 150 is adulterated today.
Canada: Global Leader in Ethyl Carbamate Regulation
This is one of the few corners of food law where Canada isn’t choosing to harmonize with anyone. In 1985, Canadian authorities detected high levels of ethyl carbamate in alcoholic beverages, mainly in imported spirits, and responded with the first limits of their kind: 30 ppb for table wine, 100 for fortified wine, 150 for distilled spirits, 400 for fruit brandies and liqueurs, and 200 for sake. The WHO’s IARC monograph records that many other countries then adopted the Canadian guidelines. The United States never set a binding limit, accepting voluntary industry commitments instead; the European Union has none either; Brazil set 150 ppb for cachaça in 2005 and moved it up to 210 in 2014.
For four decades, Canadian numbers were, in a sense, the global standard, and they sat on a somewhat discretionary administrative list. To see how that worked in practice, look at who was doing the buying.
How Guidance Numbers Become Operational
As a matter of policy, LCBO tests virtually every product it sells. New listings are tested on receipt, the board samples again annually, and ethyl carbamate is on the analytical panel for wine, cider, beer, sake and spirits. A failure escalates from warehouse hold to retail withdrawal to, if necessary, public recall, with the supplier carrying the costs.
The board's published chemical guidelines, first issued in 1998 and last revised in February 2013, set board maximums numerically equal to Health Canada's administrative values, and then went further than the federal list. For example, beer, which has never had a federal ethyl carbamate limit, gets 15 µg/L from the LCBO, or 30 µg/L above 8.5 percent alcohol.
Those units deserve a pause, because they are not the federal units. The LCBO states its thresholds in micrograms per litre, a volume basis. The federal list states parts per billion, which is parts per billion by weight; the administrative entries it replaced were expressed in micrograms per kilogram. The two numbers are numerically equal but they’re different depending on the density of a product. So, for a spirit at 40 percent alcohol, with a density near 0.95, the board's 150 µg/L corresponds to about 158 µg/kg, making the federal maximum the marginally stricter of the two. The margin is small but not zero, and the point is that a specification treating the two as interchangeable is wrong in a way a laboratory will eventually notice.
Then there is LCBO’s acceptance rule: pre-shipment samples are accepted only if the reading comes in below 85 percent of the guideline, which for spirits means 127.5 µg/L rather than 150.
The same document runs in the other direction too. Wines over $30 a bottle in quantities under 200 cases a year get 150 µg/L against a 30 µg/L guideline. Rare sherries over $70 a case wholesale, in the same small volumes, get 400. Sake at 14 percent alcohol or more gets 200. All three are marked "under review by Health Canada." So, in the old days before July 24, 2026, the LCBO was able to use a guidance number and built its own tolerances.
It also built them case by case. In November 2015, whisky author and friend of the firm Mark Bylok reported at Whisky Buzz that Sazerac 18, an aged American rye, had failed LCBO lab testing. Ethyl carbamate was the reason given to him and to others who called the board's information line, though a spokesperson would not confirm it, on the basis that results are proprietary. Then, the board told him that the LCBO overturned its original decision after reviewing the results, and, in Bylok's words, "with rare wines and whiskies, the LCBO will make an exception due to the limited distribution of the product." Twenty-four bottles reached Ontario and, in the old days, rarity and small volume could buy accommodation.
Quebec provides a helpful example because it’s on the record in a similar issue. In February 2010, a La Presse columnist asked the SAQ why Rhum Barbancourt, the Haitian cane-juice rum, had vanished from Quebec. The answer from the SAQ's communications directorate: a Health Canada directive sets the acceptable concentration at 150 ppb, the Barbancourt ambré and Réserve spéciale exceed it, and so the SAQ cannot sell them: "Nous ne pouvons donc pas déroger" or, “we cannot derogate from it.” A guidance document, non-binding by definition, described by the buyer as something from which no derogation was possible. Barbancourt returned to SAQ shelves in June 2011 after eleven years, following a process modification, and at that time only the younger expressions came back. The eight-year Réserve Spéciale is on SAQ shelves today.
For years I read the two boards as differing in appetite for enforcement, but the 2010 Barbancourt example corrects me: the SAQ was enforcing the same 150 ppb, in writing, in the language of no derogation, and Barbancourt spent eleven years off its shelves for exceeding it. What the two stories actually show is that the number never needed to be law to be decisive. A producer that failed the guidance limit typically lost access to the market.
This is part of why Canadian retail surveys look clean. Across two multi-year cycles between 2016 and 2022, the CFIA tested 615 alcoholic beverages; among the 479 subject to a maximum level, compliance was 99.8 percent. The one exception is instructive: an imported fortified plum wine at 14 percent alcohol tested at 127 ppb against its 100 ppb maximum. Health Canada assessed the levels observed as not expected to pose a concern to human health, and there was no recall. Under a guidance regime, that was where the story ended: an assessment and a judgment call. Since July 24th, the same result makes the bottle adulterated on the face of the Regulations, regardless of whether anyone thinks the product would hurt you.
Are Cane Spirits Unviable in Canada?
Ethyl carbamate presents two problems.
In wine and sake, a principal precursor is urea, which yeast produce from arginine and excrete into the ferment, where it reacts with ethanol. That pathway has largely been engineered away through low-urea yeast strains, nitrogen management, and the enzyme acid urease, which destroys residual urea before it can react. Health Canada permitted urease in 2013 for wine and sake specifically, in its own words “to reduce ethyl carbamate formation.” It works, and across the targeted categories CFIA has found high compliance levels per their testing, which is why these numbers could become binding without much drama.
Cane spirits add a second pathway. Sugar cane carries its own cyanogenic glycoside, and the cyanide released oxidizes to cyanate, which reacts with ethanol. Early cachaça research found urea contributing little in cane spirit, and later experimental work found that urea supplementation stimulated ethyl carbamate formation there too. It’s not that urea is irrelevant to cane spirit, but the cyanide pathway is often more important, which means that the wine-and-sake fix doesn’t solve the problem. Formation can also continue after distillation: copper catalyzes the cyanide-to-cyanate step, light can promote it, and it tends to increase through barrel aging.
Then there is the French appellation. The INAO cahier des charges for Rhum de la Martinique mandates fermentation in open vats and distillation on a single continuous column, with specified components made only of copper, and it prohibits rectification outright. A producer can still live within the appellation while working cane quality, fermentation management, precursor control, column operation within the specification, storage, and packaging that keeps light out. But double distillation, which the cachaça work found cuts contaminant levels, cannot happen on a single continuous column, and rectification is banned by name.
What the French trade press has documented is what producers do in practice, which is select: houses picking through casks for the ones that test low enough for North America, gaps in ranges, some leaving the market. Unpredictability is its own deterrent and an importer who cannot price the likelihood of a failed shipment inevitably will stop trying.
Bringing this back to the LCBO's 85 percent rule: the guideline states the rule without giving a rationale, but its effect is headroom against a concentration that keeps changing after the certificate is signed. A spirit accepted at 125 under the pre-shipment rule can be over 150 by the time it reaches a shelf in Etobicoke or the consumer's liquor cabinet. And this is the importers' coin flip from the top of this post: rhum agricole inherently lives near the line, is made under rules that keep it there, and is sold against a number that isn't static.
What we Learned from the Consultation
Health Canada's published summary of its 2025 consultation identified aggregate comments from alcoholic beverage manufacturers, industry representatives, a provincial government and international partners. So, who said what is not on the record. But the aggregate comments along with Health Canada’s surveillance data changed the proposal. Fortified-wine comments pointed out that these products are labelled by age rather than vintage year, and the proposed "vintage year earlier than 1995" test became "aged 20 years or greater," which, as Health Canada notes, expands the covered products from those roughly thirty years old to those twenty years old. Rice-beverage comments came with data showing beverages other than sake pattern like sake, and the category became all alcoholic beverages from rice at sake's 200; maple-sap comments came with data and produced a 200 ppb listing that did not exist before. The baijiu submission asked for 400 and supplied no supporting data. It nonetheless prompted Health Canada to revisit its own surveillance and the literature, which put roughly 94 percent of baijiu samples below 200 ppb, where baijiu ended up. Even rebuffed concerns are informative from a regulatory development perspective: comments questioned whether some whiskies could meet 150, and Health Canada answered from its own results that they can.
The published account records no comment concerning rum, cane spirit or rhum agricole. This doesn’t mean that no cane spirit stakeholders wrote in, and it certainly does not establish that 150 is the wrong number for cane spirits. But it appears as though few if any cane spirit stakeholders brought any evidence that was considered, and thus cane spirits stayed in the proposed 150 ppb band.
There’s an obvious lesson here for stakeholders involved with any regulated good, which is that responding with evidence-based comments to Health Canada and CFIA consultations is a really important habit. But there’s a less obvious lesson: the comment window is not the only door. The neat thing about Health Canada's stated program is to consolidate the administrative maximum levels into the incorporated list, which involves reviewing each substance as it transfers. And the List itself says Health Canada may modify it as new scientific information becomes available. So, with some coordinated work, it’s possible to re-open this issue. It’s just far easier to shape outcomes like this NOM while a number is still administrative than after it is referenced in the Regulations.
What to Do
Its possible that some beverage alcohol producers and importers are affected by this change in regulation. If you work with products that are inherently higher in ethyl carbamate, it may be wise to do a few things now that this change has been implemented:
Because there is no transition period and no grandfathering, find out where your products sit. A listed product over its maximum level is now considered adulterated, including stock already in the channel. If you have reason to think a SKU is over, that may be a hold-and-assess situation rather than a note for the next shipment. Depending on the facts, it may warrant quarantine, testing, and seeking advice on withdrawal, recall and whether to approach the board and the regulator before someone else does.
When testing, consider sampling down the supply chain rather than relying on the distiller. Representative finished-package samples, validation across shelf life, lot-to-lot variation and the analytical uncertainty of the method all bear on whether you can say a product complies. As discussed above, the federal maximum is by weight, so a certificate reported per litre needs converting before anyone declares a margin. Also a gentle reminder regarding the 85 percent rule in Ontario.
Think about where your SKUs sit against the five categories in the NOM and write down why. The old category names are now gone and we received a surprise interpretation from Health Canada. A good warranty would be for supplier specifications to cite the category or paragraph and the number rather than "compliant with Canadian requirements." Maximum levels of a food used as an ingredient in another food are expressly captured here, so, the limit attaches to an ingredient beverage before assembly.
A provincial liquor board can waive its own purchasing standard, because the standard is its own. But it cannot waive adulteration, so case-by-case admissions of product over the federal maximum, like the one reported behind those 24 bottles of rye, are no longer available. Stricter purchasing thresholds remain a board's prerogative, board standards for products with no federal maximum, like beer, are untouched, and some old exceptions map onto a federal category in any event (a rare sherry aged twenty years or more sits at 400 ppb under paragraph (c) regardless). Where a board document still recites a tolerance above the applicable federal maximum, expect it to be revisited, and do not build a listing strategy on it.
If your category has a problem with an administrative maximum level, get involved. The consultation record above shows that categories whose stakeholders brought data were reconsidered. Developing good evidence can re-open a proposal. And we’ll close on a friendly reminder to those who track these sorts of things that proposals surface on Health Canada's notices of proposal and notices of modification page and not in the Canada Gazette.
---
Glenford Jameson is the principal of GSJ&Co., a Canadian food law and regulatory affairs practice. This post is for informational purposes only and does not constitute legal advice.