Do balms need preservatives? Water activity, and the exceptions that matter

A true anhydrous balm needs no preservative: microbes cannot grow without free water. The mechanism, the water activity thresholds, and what flips the answer to yes.

Almost every argument about preserving a balm is really an argument about water. Microbes need liquid water to grow, and a balm made only from oils, butters and waxes has no aqueous phase for them to live in. That settles the ordinary case. The interesting work is in the exceptions, and in what the standards actually say.

What a preservative is actually for

A cosmetic preservative kills or suppresses micro-organisms inside the product. It works in the water phase, because that is where microbes live: bacteria, yeasts and moulds cannot metabolise, divide or move without liquid water available to them.

This is why anhydrous formulation defines the whole balm category. A lotion is 60 to 80 percent water and a rich growth medium that must be preserved and tested. A balm is a suspension of solids in oil. Ask what a preservative would dissolve into, and in a correctly made balm the answer is nothing.

Water activity, and the numbers microbes need

The technical measure is water activity, written aw. It is not water content. It is the ratio of the vapour pressure of water in the material to that of pure water at the same temperature, on a scale from 0 to 1 where pure water is 1.0. What matters is how much of the water is available, not how much is present.

Three points there are routinely misreported. S. aureus is the only organism in the set that grows below 0.90, and the single figure of 0.86 that circulates online is imprecise: 0.86 is an aerobic minimum, anaerobic growth is put at 0.90 to 0.91 depending on source, and toxin formation needs 0.85 to 0.88. Gram negative bacteria are generally more sensitive to low water activity than Gram positives. And spores are a separate question from growth: they germinate down to about 0.80.

What ISO 29621 actually says

The cosmetic standard here is ISO 29621. It lists six physical and chemical factors that can make a product microbiologically low risk, and a product qualifies on one of them, or on a combination.

Meeting a criterion means a challenge test to ISO 11930 is generally unnecessary. It does not remove the expectation of enumeration testing to ISO 17516, because water activity does not correlate with microbial load: a product can be bone dry and carry a high count. Any exemption must rest on justification and reliable data, not on the category name.

Two nuances are almost never published, and both matter to a balm maker.

1. 0.75 is a pragmatic industry line, not a biological one. The same standard cites 0.6 as the approximate minimum for growth of selected organisms, and some moulds multiply between the two. Industry works to 0.75 because those extreme organisms are uncommon in a cosmetic environment, not because nothing lives below it.
2. Hot filling should not be leaned on by itself. A German expert working group, with members from Beiersdorf, the BAV Institut, Mibelle and WALA, holds that water activity alone can qualify a product but a high filling temperature alone should not, because hot filling reduces the count at that moment and does nothing about recontamination afterwards. Note which criterion that is: filling above 65 C, whose worked example in the standard is lip balm and lipstick, is exactly what a balm maker would instinctively claim, and it is the weakest entry in the table. Treat it as a second factor behind water activity, never the first. Filling ranges are on pour temperatures.

How dry is "anhydrous", measured rather than assumed

Balm writing, this page included until now, asserts that an anhydrous product sits near a water activity of 0.2. It is not reliably true.

An oil carrying under a tenth of one percent water reads 0.75, the ISO 29621 line, and a fraction more takes it to 0.93. "No added water" therefore does not automatically mean a water activity near 0.2. Oils are mildly hygroscopic and equilibrate toward the relative humidity around them: equilibrium relative humidity in percent is water activity times 100. That is the real reason keeping a balm away from steam and humidity is more than housekeeping folklore.

Do not overstate it either. Water availability measured in a lipid continuum is not a bulk aqueous phase for a cell to inhabit, and a balm has no bulk aqueous phase. No source we have found reconciles those two readings of the same number, and no measured water activity values exist in the published literature for shea butter, cocoa butter, beeswax or candelilla wax at all, so anyone quoting one for a balm is estimating. The defensible position: a plain balm has no aqueous phase and will not support growth, which describes the mechanism better than a tidy 0.2, and humidity control is part of that mechanism rather than an optional extra.

Preservative, antioxidant, chelator: three different jobs

This is the most common confusion in online balm advice. These three classes of additive solve three unrelated problems, and none substitutes for another.

So adding vitamin E to a balm protects the oils and does nothing for microbial safety. Add tocopherol to a balm containing honey or aloe and you have an unpreserved water-containing product with fresher oils in it. The law agrees: EU Regulation 1223/2009 defines a preservative at Article 2(1)(l) as a substance intended exclusively or mainly to inhibit micro-organisms in the product, only Annex V substances may be used as one, and tocopherol is not on Annex V.

Inputs that bring water into a balm

Read your ingredient list and ask of each line: did this arrive with water attached? Extracts are the trap, sold in oil, glycerine, water or alcohol under similar names.

Water that arrives after you fill the jar

The formula is only half the question. A wide-mouth jar dipped into with wet fingers takes a small water inoculation on every use, and that water sits on a surface that will not disperse it. Four patterns change the answer: bathroom storage, a communal jar, a product deliberately used with water, which is the premise of a cleansing balm, and any jar contaminated with something nutritious, meaning fingers, hair or an open rim.

What the authorities prescribe is not "add a preservative". It is packaging and use instructions. The expert working group's rule is that products preserved by reduced water activity alone require waterproof primary packaging, especially where the product is likely to be used near water. FDA frames the other half as consumer behaviour: do not add water or saliva to a cosmetic, and finger-dipping is a listed contamination route. So the control is a twist-up tube, a squeeze tube or an airless pump, and where a jar is unavoidable, a spatula in the box. The packaging guide compares formats.

A decision tree you can run on your own formula

Work through it in order and stop at the first yes.

1. Does any input contain water? If yes, you need a preservative and an emulsifier, and you are no longer making a balm. Reformulate, or commit to an emulsion properly.
2. Did any botanical go in without being fully dried? If yes, discard the infusion. Residual moisture in an infused oil is the classic route to mould in a balm.
3. Will water routinely be introduced in use? Shower, wet fingers, shared jar. If yes, change the pack to a tube or pump before you reach for a preservative.
4. Is there anything hygroscopic at a meaningful percentage? Glycerine, honey powder, sodium lactate, some clays. If yes, keep the total low, use a tight closure, and be conservative on shelf life.
5. None of the above? No preservative. Add an antioxidant if the oil blend is unsaturated, keep headspace low, and manage rancidity and oxidation instead, which is the failure mode you will actually meet.

Testing a doubtful product properly

If you ran the tree and are still unsure, the answer is not a home experiment. The formal test is a challenge test to ISO 11930 or USP chapter 51: a laboratory inoculates the product with a defined panel (S. aureus, Pseudomonas aeruginosa, E. coli, Candida albicans, Aspergillus brasiliensis) and counts survivors over 28 days.

Send 50 to 100 g of the exact finished formula in its intended packaging. Plan on four to six weeks and a cost in the low hundreds per formula, as a planning figure rather than a quotation. In the UK or EU the result feeds the safety assessment and product information file, covered on selling balms in the UK and EU. Enumeration to ISO 17516 is a separate, cheaper test, and it is the one that still applies to a low-risk product.

Home tests are not conclusive. A supermarket agar plate detects only what grows readily on it, and the organisms that matter most for safety are not the ones that make a product look or smell wrong. Use home ageing for stability, texture and rancidity, which is what shelf life testing covers, and a laboratory for microbiology.

What honest labelling looks like

An unpreserved anhydrous balm should not carry a period-after-opening symbol implying microbiological protection it does not have. It should carry a realistic best-before date driven by the oxidative life of the oils, plus storage guidance that names humidity alongside heat and light. The conventions are unpacked in reading a balm label. Your own hygiene is the other half, since a rinsed tin still damp inside has introduced the water the formulation avoids. See workshop hygiene.

Where to go from here

If your formula passed the tree, stop worrying about preservatives and start worrying about oxidation, which is what will end the product's life. Keep tins full, cool and dry, and date every batch. If it failed, the useful move is almost never a natural preservative. It is to take the water out, change the packaging, or accept that you are formulating an emulsion and do that properly.

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