Nut free balm: which balm ingredients are actually tree nuts, and which are not
A nut free balm at 9 percent wax with a botany table separating true tree nuts from drupes, substitution swaps, a 30 ml tin, a 64 to 70 C pour and labelling rules.
Most of the ingredients people take out of a balm to make it nut free are not tree nuts, and one or two of the ones they leave in are. This page sorts the botany from the allergen law, gives a formula that carries no nut derived material at all, and is honest about the part that matters most: what a maker can truthfully put on the label at the end of it.
Almond, macadamia and hazelnut are listed tree nut allergens. Shea, argan, coconut and apricot kernel are not. A working nut free tin balm is 9 percent beeswax, 18 percent kokum butter, 45 percent high-oleic sunflower oil, 27.9 percent rice bran oil and 0.1 percent mixed tocopherols. Melt at 75 to 80 C, pour tins at 64 to 70 C, cure 48 hours.
Which balm ingredients are actually tree nuts
Two lists decide what counts, and neither is botanical. Annex II of the EU food information regulation names eight tree nuts: almond, hazelnut, walnut, cashew, pecan, Brazil nut, pistachio and macadamia. FDA's list runs slightly longer and was narrowed in 2023 draft guidance, which left coconut and shea nut off it. Everything else in the balm cupboard is a seed, a fruit kernel or a drupe, whatever the supplier's catalogue calls it.
| Ingredient | Botany | Listed tree nut | Position in a nut free formula |
|---|---|---|---|
| Sweet almond oil | Rosaceae, the seed of a drupe | Yes | Out. The one most often in a balm and the one to remove first |
| Macadamia oil | Proteaceae, a true nut | Yes | Out, although reported allergy is genuinely uncommon |
| Hazelnut oil | Betulaceae, a true nut | Yes | Out, and it carries birch pollen cross-reactivity as well |
| Shea butter | Sapotaceae, a fruit kernel | No | Not a tree nut, but see the note below on why it still comes out |
| Argan oil | Sapotaceae, a fruit kernel | No | Not a tree nut. Marketed as a nut oil, which is the whole problem |
| Coconut oil | Arecaceae, the drupe of a palm | No | Its own argument, below |
| Apricot and peach kernel oil | Rosaceae, same genus as almond | No | Unlisted, but the closest relative of almond on the shelf |
| Sesame oil | Pedaliaceae, a seed | Separately listed, not a nut | Out of this formula. A rising allergen and usually unrefined |
| Peanut oil | Fabaceae, a legume | Separately listed, not a nut | Rare in balms. Capped at 0.5 ppm peanut protein in EU and UK cosmetics |
| Oat oil and colloidal oatmeal | Poaceae, a cereal grain | No | Not a nut. It is a gluten question, not an allergen list question |
Shea gets its own answer because it is genuinely two-part. Legally and botanically it is not a tree nut: it is the kernel of a fruit of Vitellaria paradoxa, it appears on neither list, and testing of commercial shea for IgE binding soluble proteins has come back essentially empty. It comes out of this formula anyway, because a customer asking for nut free is not asking a botanical question and a butter universally called shea nut butter cannot be argued into acceptability at a market stall. That is a commercial decision and should be described as one rather than dressed up as a safety finding. The material itself is covered in shea butter.
Coconut is awkward from the other direction. It is a palm drupe, coconut allergy is uncommon, and it does not reliably track tree nut allergy, but some older American guidance treated it as a tree nut and a fraction of customers still read it as one. There is a formulation argument for leaving lauric fats out of a lip balm as well, independent of the allergen question.
What refining does to the protein, and where the guidance stops
The allergens in nuts are storage proteins, not fats. They are water soluble and they partition into the press cake, so a pressed oil starts with very little and full refining removes most of what remains. Alkali neutralisation carries protein into the soapstock, bleaching clay adsorbs polar material and deodorisation steam strips at 180 to 270 C. The measured gap is large: refined oils have been reported at roughly 0.2 to 2.2 micrograms of protein per mL against 100 to 300 for crude and cold pressed grades.
Three caveats travel with those numbers, and dropping any of them turns a reasonable statement into a false one. First, almost all of the good data is peanut: allergic volunteers challenged with refined peanut oil did not react where a minority reacted to crude, and nobody has run that trial on almond. Second, the food law exemptions are narrow and inconsistent, since US law excludes highly refined oils from the definition of a major food allergen while EU and UK food law exempts only fully refined soybean oil. Third, none of it is about skin: challenge data describe a gut exposure to milligrams of protein, and a balm on intact skin is a different route and a different dose.
What that adds up to is narrow but usable. Refined nut oils very probably present a low risk, you cannot promise that to anyone, and the judgement belongs to the customer's allergy clinician. The evidence, the residue table and the labelling position are in nut oil allergy in balms. This page takes the other route: build the product so the conversation never has to happen.
Replacing shea means replacing two things, not one
This is where nut free reformulations usually go wrong, because shea does two separate jobs and most substitution tables address only the first.
The first is solid fat: shea is roughly 30 to 45 percent stearic acid and melts across 32 to 38 C, so it gives hardness and body. The second is cushion, which comes from a fatty acid spread wide enough that the butter softens gradually rather than all at once, plus an unusually large unsaponifiable fraction of 5 to 17 percent. That is why a shea balm feels rich at a fingertip and keeps feeling rich as it warms.
Kokum does the first job better than shea and the second hardly at all. It is 50 to 60 percent stearic, roughly 70 to 80 percent of its triglycerides are the single species SOS, and it melts across a narrow 38 to 42 C band with only 1 to 2 percent unsaponifiables. So it is harder, sets crisply, almost never grains, and reads dry and fast rather than cushioned. Swap it in one for one and the balm comes out firmer and thinner feeling at once, which is the complaint people describe as "it went waxy".
| Job shea was doing | The number behind it | Replaced here by | What changes |
|---|---|---|---|
| Hardness and body | 30-45% stearic, melts 32-38 C | Kokum butter at 18% | Firmer set, sharper melt at 38-42 C |
| Cushion and slow melt | 5-17% unsaponifiables | Rice bran oil at 27.9% | Cushion returns as slip rather than as body |
| Bulk and spread | 40-55% oleic | High-oleic sunflower at 45% | Lighter, faster, essentially neutral in odour |
| Oxidative life | 1-2 years typical | Both oils are high oleic or stable | Comparable, with tocopherol at 0.1% |
Rice bran oil is the piece that makes this work. At roughly 38 to 42 percent oleic and 32 to 38 percent linoleic it is heavier and slower on the skin than sunflower without being greasy, and it carries gamma-oryzanol and a large tocopherol and tocotrienol fraction. It is a grass seed oil with no nut association at all, and the closest thing on the shelf to cushion that does not come from a butter. Compare it in carrier oils compared and against the solid fats in butters compared.
Eighteen percent kokum is above the 5 to 15 percent band usually quoted for it, and deliberately so. That band assumes kokum is an add-on hardener working alongside another butter. Here it is the only butter in the formula, so it has to carry the whole solid fat load. Usage rates and the shea conversion arithmetic are in mango and kokum butter.
The substitutions, ingredient by ingredient
| Take out | What it was doing | Put in | What to expect |
|---|---|---|---|
| Sweet almond oil | Light, cheap bulk emollient | High-oleic sunflower oil, one for one | Almost indistinguishable, and longer shelf life |
| Macadamia oil | Cushioned, skin-like feel from palmitoleic acid | Rice bran oil, or squalane at 5 to 10% | Rice bran is heavier; squalane is drier and dearer |
| Apricot or peach kernel oil | Light, fast, faintly silky | High-oleic sunflower or high-oleic safflower | Very close. Chosen mainly to avoid the Prunus association |
| Argan oil | Marketing, plus a good tocopherol content | Rice bran oil | Similar unsaponifiable interest at a fraction of the price |
| Shea butter | Hardness and cushion together | Kokum butter plus rice bran oil | Firmer, drier, far less prone to graining |
| Cocoa butter | Snap and wear | Kokum butter at about 0.8 times the weight | Similar snap, no chocolate odour, no bloom |
| Coconut oil | Cheap hardness, fast absorption | Kokum, or caprylic/capric triglyceride for the light feel | Less occlusive, less likely to feel drying on lips |
| Sesame oil | Traditional massage carrier | Rice bran or high-oleic sunflower | Sesame is usually cold pressed, so this is a real reduction |
| Colloidal oatmeal | Claimed soothing, and a gluten question | Leave it out, or use kaolin for slip | Oat is not a nut. Remove it only if oat free is the actual brief |
Two things stay. Beeswax is an insect product and carries no nut question at all, although it has its own contact allergy literature. Mixed tocopherols are usually derived from soybean or sunflower distillate, so ask which: a soy derived tocopherol in a product sold as allergen conscious is worth knowing about, even though refined soybean oil is exempted in both food regimes.
The formula
Weight percent of the total batch, adding to exactly 100. Gram columns are for a 100 g test batch and a 250 g production batch, which fills eight 30 ml tins at a 27 g fill after 10 percent overage for what stays on the jug and spatula.
| Phase and ingredient | % | 100 g (g) | 250 g (g) | What it does |
|---|---|---|---|---|
| A: beeswax, cosmetic grade | 9.0 | 9.0 | 22.5 | Just enough structure to stop slump in a tin |
| A: kokum butter | 18.0 | 18.0 | 45.0 | The whole solid fat load. Hardness and clean set |
| A: high-oleic sunflower oil | 45.0 | 45.0 | 112.5 | Bulk emollient, neutral, stable, cheap |
| A: rice bran oil | 27.9 | 27.9 | 69.75 | Cushion and slow break. Replaces shea's feel |
| B: mixed tocopherols | 0.1 | 0.1 | 0.25 | Antioxidant, added below 60 C |
| Total | 100.0 | 100.0 | 250.0 | Wax 9, butter 18, liquid oils 72.9 |
At 100 g the tocopherol is 0.1 g, which is at the edge of what a domestic scale reads honestly, so make the 250 g batch or leave it out and accept a shorter life. The working range is 0.05 to 0.2 percent of a mixed tocopherol concentrate and more is not better, for the reasons in vitamin E and antioxidants. For a lip version, take beeswax to 12 percent and kokum to 15.
Method, pour and cure
- Weigh phase A into one jug. Beeswax, kokum, sunflower and rice bran, tared between each addition. Weigh the tocopherol separately.
- Melt at 75 to 80 C. Stand the jug in simmering water and hold until the liquid is completely clear. Kokum melts at 38 to 42 C and dissolves easily, but the beeswax needs the temperature.
- Cool below 60 C and add phase B. Stir in the tocopherol off the heat. There is no flavour or fragrance in this formula, which is deliberate: an allergen conscious product is a poor place to add fragrance allergens.
- Pour tins at 64 to 70 C. Fill each 30 ml tin in one movement to about 2 mm below the rim. Hotter than 72 C and you get a deep shrinkage dip; cooler than 62 C and the surface sets rough.
- Cool uncovered and undisturbed. Leave the tins open at room temperature until firm, roughly an hour. Lidding them warm traps condensate on the underside of the lid, which is one of the few ways water gets into an anhydrous product.
- Cure 48 hours before judging. Kokum crystallises fast into a single dominant lattice, so the balm reads hard at two hours and settles noticeably by two days. Reformulating on the early reading is how people end up cutting the butter too far.
The pour window is narrow because the balm shrinks as it sets and a tin poured hot dishes in the middle. The physics, and what changes at larger batch sizes, is at pour temperatures.
Judge the finished balm on a cool day, not a warm one. Kokum's melt range sits just above skin temperature, so the balm needs a second or two of contact before it releases, and a warm kitchen will flatter it. If it still drags after two seconds at 20 C, take kokum to 15 percent and add the three points to rice bran oil.
Cross-contact: the supplier, the specification and your own jug
A formula with no nut derived ingredient is not the same as a product with no nut protein in it, and the gap between the two is almost entirely about equipment.
Start with what a supplier's specification sheet tells you: INCI name and botanical species, the refining route if you are lucky, free fatty acid and peroxide values, colour, often heavy metals and pesticide residues. It is a purity document. It does not tell you what else ran through that refinery, whether the filling line handled almond oil that morning, or whether this season's crop came from a different origin. Commodity oils such as sunflower and rice bran are routinely refined in multi-oil plants, which is normal rather than alarming but has to be asked about rather than inferred. Ask for an allergen statement separately, keep the reply on file with the batch, and re-ask when the origin changes. What to ask, and what a reasonable answer looks like, is in sourcing ingredients.
Then your own bench, which is the larger risk by a wide margin. Pour an almond oil balm and a sunflower balm through the same jug and the second contains almond oil, not a trace but a measurable amount, because a warm jug retains a film of a gram or more. Protein residue is carried in the oil rather than dissolved in it, so a solvent rinse that clears the fat does not necessarily clear the protein. What works is hot water above 60 C, detergent, scrubbing and a rinse. Spouted jugs, silicone spatulas and filling nozzles hold the most, and scratched plastic is worse than stainless or glass. The regime is in workshop hygiene.
Sequencing beats scrubbing. Run nut free batches first in the day on clean equipment, before anything containing almond, hazelnut or macadamia is opened, and keep a dedicated jug and spatula if you can afford the duplicates. Record the running order against the batch code so you can answer a question about one specific tin months later, which is what batch records are for.
What goes wrong
| Symptom | Mechanism | Fix |
|---|---|---|
| Harder than the shea version it replaced | Kokum is more stearic and melts higher, and 18 percent is above the equal-hardness swap | Kokum to 15 percent, the balance into rice bran oil. See too hard or too soft |
| Drags rather than glides on first contact | Kokum's 38 to 42 C melt sits above skin temperature, so the film needs a moment to release | Beeswax to 7 percent, or add 5 percent castor oil out of the sunflower. See drag and poor glide |
| Thin and oily rather than rich | Not enough solid fat once shea's unsaponifiables are gone | Kokum up, not liquid oil down. Do not exceed 22 percent |
| Beads of oil on the surface | 72.9 percent liquid oil is more than 9 percent wax can always hold in a warm room | Beeswax to 10 or 11 percent. See balm sweating |
| Smells of cardboard after eight months | Rice bran oil is a third linoleic and oxidises faster than the sunflower around it | Keep tocopherol at 0.1 percent, store cool, shorten the date. See rancid balm |
| A batch that behaves unlike the last one | Supplier changed origin or refining grade without telling you | Log the lot number, and test one tin before filling the run |
| A customer reacts anyway | The reaction was never about nuts. Fragrance, beeswax and propolis are more frequent causes | Give them the full ingredient list to take to their clinician |
What a nut free label can and cannot promise
Cosmetics are not regulated like food, and this is the most misunderstood point in the subject. There is no food style allergen declaration for nuts in cosmetics in the EU, the UK, the US, Canada or Australia. The law requires the whole formula instead: Article 19 of Regulation (EC) No 1223/2009, retained in Britain as the UK Cosmetics Regulation, requires an ingredient list in INCI nomenclature, and 21 CFR 701.3 does the same job in the US. Because INCI names are binomial the botany is already on the pack, since Prunus Amygdalus Dulcis Oil is sweet almond and Macadamia Ternifolia Seed Oil is macadamia. That is why reading a balm label is worth teaching a customer.
Absence claims are where makers get into trouble. In the EU and UK they fall under the common criteria of Commission Regulation (EU) No 655/2013, which require every claim to be truthful, evidenced and not misleading. The Commission's technical document on cosmetic claims goes further and treats "free from allergens" as a claim that cannot be permitted, on the ground that a complete absence of allergenic risk cannot be guaranteed. In the US, MoCRA added registration, listing and safety substantiation but did not import food allergen labelling into cosmetics, and "hypoallergenic" remains a term with no legal definition and no test behind it. The jurisdiction specific duties are in selling balms in the UK and EU and labelling cosmetics in the US.
So the honest label is not "nut free". It is a complete INCI list plus a voluntary plain English line such as "formulated without nut derived oils", which is a statement about your formula and can be checked against it. If you also make almond oil products, describe your control in writing when asked: dedicated jugs, nut free batches poured first, batch codes logged. Do not put a precautionary "may contain nuts" on the tin, because that is a borrowed food convention with no cosmetic equivalent and it warns without informing.
The decision rule is short. If you sell to strangers at markets, formulate the nut oils out and keep the conversation to what is in the tin, because the alternative is fielding a clinical question you are not qualified to answer. If you make for people you know, refined almond oil is very probably fine and their allergy team is the one to say so. Either way you cannot assess anyone's allergy, and telling a worried parent a product is safe for a nut allergy is outside your competence and, in Britain and Europe, an unsubstantiated claim. The limits of what this site can tell you are in the disclaimer.
Frequently asked questions
Is shea butter a tree nut?
Not for allergen purposes. Shea is the kernel of a fruit of Vitellaria paradoxa, it appears on neither the EU Annex II tree nut list nor FDA's, and testing of commercial shea for IgE binding soluble proteins has found essentially none. Many makers still remove it from a nut free product, because it is universally called shea nut butter and customers read it that way. That is a commercial choice, not a safety finding.
Can I put "nut free" on my balm label?
Not safely. In the EU and UK, claims must meet the common criteria of Regulation 655/2013, and the Commission's technical document treats free from allergen claims as unacceptable because complete absence of risk cannot be guaranteed. Say "formulated without nut derived oils" instead, which describes your formula and can be checked against your ingredient list.
What can I use instead of sweet almond oil?
High-oleic sunflower oil, weight for weight, is the closest and cheapest swap: similar lightness, similar spread, no odour and a longer shelf life because it is mostly oleic acid. If you want something a little richer, rice bran oil is heavier and slower. Both are seed and grain oils with no tree nut on either statutory allergen list.
Why is a kokum balm harder than the shea balm it replaced?
Kokum is 50 to 60 percent stearic acid against shea's 30 to 45, and it melts across a narrow 38 to 42 C band rather than shea's broader 32 to 38 C. It also lacks shea's 5 to 17 percent unsaponifiable fraction, which is where the cushion came from. Expect a firmer, drier balm, and give it 48 hours to cure before judging.
Is coconut oil safe in a nut free balm?
Botanically yes, since coconut is the drupe of a palm and appears on neither statutory tree nut list, and coconut allergy is uncommon and does not track tree nut allergy. Commercially it is less simple, because some older guidance treated it as a tree nut and some customers still read it that way. If the point of the product is reassurance, leaving it out costs you very little.
Does refined almond oil count as nut free?
No, and the label cannot say so. Refined oils carry very little protein, roughly 0.2 to 2.2 micrograms per mL in the published measurements, but almond remains a listed allergen in both the EU and the US at every refining grade, and almost all the challenge evidence is peanut rather than almond. A product containing it is a product containing a nut derived oil.
Do I need dedicated equipment to make a nut free balm?
If you also make balms with almond, hazelnut or macadamia oil, yes in practice. A warm jug retains a film of a gram or more, so the next batch through it genuinely contains the previous oil. Dedicated jugs and spatulas are the cheap answer; sequencing nut free batches first in the day on freshly washed equipment is the workable one.
Sources and further reading
- European Union, Regulation (EC) No 1223/2009 on cosmetic products, Article 19 and Annex III, EUR-Lex.
- European Commission, Regulation (EU) No 655/2013 laying down common criteria for the justification of claims used in relation to cosmetic products, EUR-Lex.
- European Commission, Technical document on cosmetic claims, Annex III on "free from" claims, Sub-Working Group on Claims.
- European Union, Regulation (EU) No 1169/2011 on the provision of food information to consumers, Annex II, EUR-Lex.
- US Food and Drug Administration, 21 CFR 701.3, Designation of ingredients, eCFR.
- US Food and Drug Administration, Questions and Answers Regarding Food Allergens, including the Food Allergen Labeling Requirements: Guidance for Industry, draft, Edition 5, 2023.
- Crevel R.W.R., Kerkhoff M.A.T. and Koning M.M.G., Allergenicity of refined vegetable oils, Food and Chemical Toxicology, 2000.
Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.