Comfrey salve: the pyrrolizidine alkaloid question, answered before the formula

A comfrey salve with the pyrrolizidine alkaloid limits stated in micrograms, leaf against root, a 12 percent wax formula, a 60 ml tin and a 70 to 76 C pour.

Almost every decision in balm making is a process decision: a temperature, a ratio, a cooling rate. Comfrey is the exception. The thing that makes a comfrey salve safe or unsafe is fixed before you switch anything on, in the plant part and the material you bought, and no melting, straining or filtering afterwards changes it. So this page settles the alkaloid question first and gives the formula second.

• Leaf, never root
• Leaf PA 30 to 200 mg/kg
• 100 micrograms/day external reference
• Beeswax 12%
• Intact skin only

What an oil infusion of comfrey actually contains

Comfrey's reputation rests on allantoin, and an oil infusion is close to the worst possible way of delivering it. Allantoin dissolves at around 0.5 g per 100 ml in water at 25 C and is practically insoluble in vegetable oils, so a macerate carries an unknown and small amount whatever the ratio or the steeping time. Root holds roughly 0.6 to 4.7 percent allantoin on dry weight and leaf a quarter to a half of that, and almost none of either follows the oil out of the jar.

What the oil does take up is chlorophyll, cuticular wax, triterpenes, rosmarinic acid, a little of the plant's own lipid, and a share of the alkaloids. In other words, the constituent people want stays behind and the constituent they worry about travels. That asymmetry is the whole reason this page exists, and the general mechanism is set out at herbal infused oils.

If you want a known quantity of allantoin in the tin, weigh in the purified powder at 0.2 to 1 percent and disperse it as a suspension, as described at allantoin. That is an honest ingredient at a stated level. It is not the same thing as a comfrey infusion, it must appear on the label as Allantoin, and it does not license any claim the infusion would not license.

Pyrrolizidine alkaloids by plant part

The compounds at issue are the 1,2-unsaturated pyrrolizidine alkaloids and their N-oxides: symphytine, echimidine, lycopsamine, intermedine and relatives. They are hepatotoxic, the concern is cumulative rather than acute, and they are chemically robust. That last point matters more than it sounds: they are not destroyed by the temperatures used in infusion, they are not removed by muslin, and they are not diluted by a longer steep. The number on your finished tin is decided by the herb.

An order of magnitude between root and leaf is the entire safety argument, and it holds regardless of which threshold you believe. The wider comparison with arnica, St John's wort and the rest of the salve herb cabinet is at arnica, comfrey and salve herbs.

Which limit applies, and to what

Most of the confusion in this area comes from quoting a number that belongs to a different body of law.

• Germany, medicines. The long standing reference figures for comfrey preparations are 100 micrograms of 1,2-unsaturated pyrrolizidine alkaloids and N-oxides per day for external use, with use restricted to four to six weeks in a year, and 1 microgram per day internally. Nearly every figure quoted online traces back to here.
• EU herbal medicines. The European Medicines Agency's herbal committee set a daily intake threshold for oral herbal medicinal products in the region of 1 microgram, since tightened. Registered medicines, not cosmetics.
• EU food. Maximum alkaloid levels for herbal infusions, teas, borage leaves, cumin and supplements sit in the contaminants rules, following the EFSA CONTAM panel's 2017 opinion. Food again, but it shows where regulators think the acceptable exposure sits.
• Cosmetics. Neither Regulation (EC) No 1223/2009 nor the assimilated UK Cosmetics Regulation sets a numeric pyrrolizidine alkaloid limit in its annexes. Health Canada's Cosmetic Ingredient Hotlist does restrict comfrey, so a product sold across borders can meet a limit its home market never imposed.

The absence of a cosmetic number is not permission. In the UK and EU the control is the safety assessment: a qualified assessor has to justify the exposure with a measured figure for the raw material, your inclusion level and a realistic application amount, which is why a comfrey product with no analysis behind it stalls a safety assessment and CPSR. The rest of the market obligations, from the responsible person to the notification, are at selling balms in the UK and EU. For a salve made for your own use the law asks nothing, and the arithmetic below still applies.

Sourcing: cultivar, plant part and the question to ask

Two species turn up under the name. Symphytum officinale is common comfrey. Symphytum x uplandicum, Russian comfrey, is a hybrid with prickly comfrey, and the named clones sold to gardeners, Bocking 14 among them, belong to it. Bocking 14 came out of the Bocking trial ground selections and is the usual recommendation, for reasons worth separating from each other.

The good reasons are traceability and consistency. It is a sterile clone propagated from root cuttings, so every plant is genetically identical, it does not set seed and spread, and it yields leaf abundantly enough that nobody needs to dig roots. That last point does most of the safety work, because it removes the temptation that makes salves dangerous.

The weak reason is the one most often repeated. The claim that Bocking 14 is inherently low in alkaloids comes largely from horticultural and fodder literature and is not supported by a body of published assays on cosmetic-grade material. Russian comfrey hybrids typically carry echimidine, which common comfrey largely lacks and which is among the more toxic congeners, so a cultivar name is not a substitute for a figure.

The infusion: 1:5 dried leaf, 40 C, 14 days

1. Harvest and dry mature leaf. Later season growth, not spring shoots. Dry in shade or a dehydrator below 40 C until a leaf snaps rather than folds. Water carried into the jar is the one genuine microbiological hazard in oil infusion.
2. Weigh 1 part leaf to 5 parts oil. Use a stable carrier: high-oleic sunflower is the sensible default, since a two week warm hold is hard on a polyunsaturated oil. Crumble the dry leaf before covering it, or it floats in a mat.
3. Hold at 40 C for 14 days. A yoghurt maker, a proofing box or a warm airing cupboard. Stir on day one to release the air trapped in the leaf hairs, then stir daily. Keep it dark and lidded.
4. Strain and settle. Two layers of muslin, press the marc, then leave the oil undisturbed in a tall jar for 24 to 48 hours and decant off the top, discarding the last centimetre with the fines and any trace water in it.
5. Label the jar. Herb, plant part, cultivar, ratio, carrier, date. Every infused oil looks identical by February, and plant part is the one detail you cannot afford to guess at later.

Expect to recover well under the oil you put in: dried leaf is light and absorbent and the marc holds two to three times its own weight, so an unpressed 1:5 macerate can give back barely two thirds. That changes your yield, not your alkaloid concentration, because the oil left in the marc is at the same strength as the oil you poured off. Run your own ratio through the infusion calculator before you buy leaf.

The formula

Weight percent of the total, adding to exactly 100. The 300 g batch fills five 60 ml tins at a 55 g fill, with working overage.

12 percent beeswax puts this at the firm end of a scoopable salve, which suits a product applied over a broad area of arm or leg. 12 percent shea butter supplies the cushion that a wax and liquid oil blend lacks; it also brings a grain risk, which the melt schedule below handles. The base is the same architecture as the all-purpose herbal salve, tightened for a single herb.

1. Melt phase A at 75 to 80 C. Hold until completely clear, 20 to 30 minutes, so no shea crystal memory survives. Ceiling 85 C.
2. Cool below 60 C and add the tocopherol. Stir it through off the heat.
3. Pour at 70 to 76 C. Return briefly to the bath to get back up. This sits above the usual 60 to 68 C tin window at pour temperatures because 24 percent solids thickens the melt early, and a cool pour into a small deep tin sets against the wall and builds a layered, scaly surface.
4. Chill, then top up. Twenty minutes in a fridge drives shea through crystallisation fast, which is the main defence against grain. Then pour a thin top-up layer over the shrinkage dish.
5. Cure 24 hours before judging, 48 before labelling. The salve reads softer at two hours than it will at two days.

The exposure arithmetic for this formula

Assume the worst case: every alkaloid in the herb transfers into the oil at the 1:5 ratio, the infused oil is 52.8 percent of the salve, and the calculation follows through to an applied daily dose. The German external reference figure of 100 micrograms per day is the comparison.

Read the rows against each other rather than against the threshold. Mature leaf at typical content leaves roughly a tenfold margin at a gram a day. Root at typical content is over the reference figure before anyone applies a second gram or dresses two sites. Young leaf at the top of its range and a generous application lands in the same place, which is why the harvest timing matters and why a realistic figure from how much balm to apply belongs in the calculation.

Two caveats, both in the conservative direction. Complete transfer is a deliberate overestimate, because N-oxides are usually the larger fraction in the plant and are too polar to move readily into a fatty carrier. And no robust dermal absorption fraction has been published for comfrey alkaloids, so an applied dose is not an absorbed dose. The margin is probably wider than the table suggests. It is not measured, which is exactly why the safe move is to take the margin rather than spend it.

Intact skin only, and why broken skin voids the sum

Every number above assumes an intact stratum corneum, which is a genuine rate limiting barrier to a polar alkaloid. Damaged skin does not raise the exposure by a knowable multiple; it removes the term that made the estimate meaningful at all. A dose that was defensible on a bruise-free forearm becomes unquantifiable on a graze, and there is no published fraction to substitute. General absorption behaviour across intact and compromised skin is covered at lipid penetration into skin.

This is more than a technicality for comfrey specifically, because its traditional uses point squarely at exactly the situations where skin is broken. A salve made for knocks will be reached for after a fall. Say so on the pack, in those words: intact skin only, not for use on broken skin, open wounds or mucous membranes. Keep it away from children, and leave it out of pregnancy and breastfeeding entirely, along the lines set out at balms in pregnancy. Cap the duration too: four to six weeks in a year, borrowed from the medicines position, is the cautious reading and there is nothing to gain from exceeding it.

What you can say, and what makes it a medicine

Comfrey topicals hold Traditional Herbal Registrations in the UK, which means the regulator has already decided what those claims are. Bruises, sprains, strains, joint pain, inflammation, healing and anything about bones are medicinal claims, and a product carrying them is a medicine by presentation under the Human Medicines Regulations 2012, whatever the label calls itself. The folk name knitbone is a claim in a single word, so it cannot be a product name. "Traditionally used by herbalists for" launders nothing.

What is left is genuinely cosmetic: softens and conditions the skin, a comforting massage medium, a rich anhydrous balm for dry skin on knees, elbows and shins. That is not much of a story, and if the story is what you are buying the herb for, this is the wrong ingredient. Work every line through cosmetic versus drug claims before any of it is printed. If the aim is a rub for tired legs after exercise, a warming formula built on declared actives at declared levels, as in muscle rub balm, is the more honest route.

Failure modes

When not to make this at all

The decision rule is short. Selling it, and you cannot obtain an LC-MS/MS figure for total 1,2-unsaturated alkaloids and N-oxides on the specific batch of leaf: do not make it. Making it for yourself, and all you can get is unidentified "comfrey herb" of unknown plant part and cultivar: do not make it. Aiming it at anyone pregnant, breastfeeding, under sixteen, living with liver disease, or with skin that is currently broken: do not make it.

If all of those clear, you have a firm green salve that behaves like any other 12 percent wax anhydrous product and carries a small, calculable alkaloid load with a wide margin. Be clear about what the herb contributes to it, which is colour, a characteristic smell and a share of lipophilic plant chemistry, not a dose of anything. The occlusive performance you can feel comes from the wax, the butter and the oil, and it would be there if you left the comfrey out.

Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.