How to make lip balm: a formula that works, and why every number is there

A lip balm formula in weight percent with three variants, exact melt and pour temperatures, tube filling technique, cure time and a full failure-mode table.

Most lip balm recipes are a ratio and a hope: one part wax, two parts butter, three parts oil, in tablespoons, and no explanation of why the result crumbles, sweats or slides off. This page gives a formula in weight percent, the reasoning behind every number, and what goes wrong at each step, so you can diagnose a batch instead of starting again.

What a lip balm actually has to do

Lip skin has no sebaceous glands to speak of and a thin, poorly organised stratum corneum, so it loses water faster than the skin a centimetre away and cannot replace the lost surface lipid on its own. A balm is a physical patch. The wax makes the film stay put, the butters give it substance and slow melt, the liquid oils carry it and let it spread. Nothing in a balm repairs anything. It reduces water loss while it is present, which is enough for chapping to settle on its own.

That gives three jobs: hold shape in a pocket at 30 C, release a film under light pressure, and stay on for an hour rather than five minutes. Wax handles the first, oils the second, butters and castor the third.

The base formula: classic beeswax tube balm

All percentages below are weight percent of the total batch and add to exactly 100. Weigh everything. Do not measure by volume.

Why each number is what it is

16% beeswax sits in the middle of the 10 to 20% window for a twist-up tube. Below 12% the stick softens in a warm pocket and drags a wet smear across the lip; above 20% it goes hard, waxy and reluctant to release, which readers usually describe as the balm not doing anything. Beeswax melts at 61 to 65 C, so the finished stick is comfortably solid at body temperature but yields under the pressure of an application.

24% total butters stops the balm feeling thin. Cocoa at 10% adds snap and wear (its profile is roughly a third stearic); shea at 14% adds the cushion cocoa lacks. Splitting them is deliberate: all-shea grains easily, all-cocoa feels chalky.

10% castor oil is the gloss and cling lever. It is 85 to 90% ricinoleic acid, which is why it is sticky; 5 to 15% is the useful window and past about 25% it turns tacky. Jojoba at 15% is not really an oil but a liquid wax ester, extremely stable, and it buys the blend shelf life. High-oleic sunflower is the workhorse: neutral, cheap, one to two years of stability, and it keeps the cost sane at scale. If you want to change proportions rather than materials, the arithmetic is set out in lip balm ratios.

Two more variants

Vegan candelilla tube balm

Candelilla is harder than beeswax at equal weight, so the substitution is roughly 0.6 to 0.7 times the beeswax figure. 16% beeswax becomes about 10% candelilla. Candelilla also feels drier, so castor goes up to restore cushion. Weight percent, totalling 100.

The full conversion method, including what else in a natural balm is not vegan, is on the vegan lip balm page.

Softer tin balm

A pot balm is picked up with a fingertip, so it must be soft enough to dent under light pressure. Drop the wax to 9% and let the oils rise. Weight percent, totalling 100.

Batch size and how many containers it fills

Work back from the container. Standard oval twist-up tubes take 4.2 to 4.9 g, so plan at 4.5 g. Add 10% overage for what stays on the jug wall, the spatula and the thermometer, and the arithmetic becomes: usable product equals batch weight divided by 1.10.

To convert a container's stated volume into a fill weight, multiply millilitres by 0.92, the working density of a set balm. Molten balm is less dense than set balm, which is exactly why a tin that looked full when you poured it has a dished centre an hour later.

Equipment: what you actually need

Three things are non-negotiable. A digital scale that reads to 0.01 g, because a 25 g batch has ingredients at 0.13 g. A thermometer, probe or infrared, because every fault below is a temperature fault. And a pouring vessel with a real spout: a stainless jug or a heatproof glass measuring jug beats a bowl and a spoon by a wide margin.

Useful but not essential: a silicone spatula, a tube filling tray, and a second small jug for the top-up pour. Sold as necessary but rarely so: dedicated double boilers (a saucepan of simmering water and a jug does the same job), pipettes for filling tubes (slow, and they clog), induction melters below production scale, and "lip balm bases" sold pre-mixed, which mostly remove the part you are trying to learn. There is more on this in the equipment guide.

Method

1. Weigh phase A. Beeswax, cocoa butter, shea butter, castor, jojoba and sunflower into one jug, tared between each addition. Weigh phase B separately into a small dish.
2. Melt at 75 to 80 C. Stand the jug in simmering water. Hold at 75 to 80 C (167 to 176 F) until the liquid is completely clear with no shea crystals. Do not exceed 85 C for long: it darkens beeswax and degrades tocopherol.
3. Cool below 60 C. Lift the jug out and stir. When it reads under 60 C (140 F), stir in the tocopherol and flavour. Anything volatile added hotter is partly lost to the air.
4. Bring to pour temperature. Tubes pour at 65 to 72 C (149 to 162 F), tins at 60 to 68 C (140 to 154 F). If you have cooled past that, return to the bath for 30 seconds rather than reheating hard.
5. Pour in one movement. Fill each tube to about 1 mm below the rim without stopping mid-tube, which leaves a visible seam. Work in rows so nothing sits and thickens.
6. Wait, then top up. After 10 to 15 minutes the surface has set and the centre has dished. Pour a second thin layer over each tube to give a level, slightly domed top.
7. Cool upright and cure. Two hours at room temperature, upright, uncapped until firm. Then cap and leave 24 to 48 hours before judging the texture.

The pour window is narrow for a reason, and the physics of it is set out at pour temperatures.

Filling tubes, the top-up pour and cure

Twist the mechanism to the bottom before filling, or you lose part of the fill inside the barrel. At 65 to 72 C the balm skins over in a couple of minutes and the dip that follows is shallow enough for one top-up to fix. At 80 C you get a deep crater and often a hollow core, which is tunnelling. At 58 C you get lumps and a rough, matte surface.

The second pour is not optional if you care how the product looks. Do it once the first pour has set enough to hold a skin, and keep the top-up jug warm so it flows over the existing surface rather than sitting on it as a separate disc. Technique detail, including jigs and multi-tube trays, is at filling lip balm tubes.

Cure matters because wax crystallisation continues for a day or two after the balm is visibly solid. A stick judged at two hours reads softer than the same stick at 48 hours, and reformulating on that early reading is how people end up with a balm that is too hard.

How to adjust it

• Harder: raise beeswax by 2 percentage points and drop the sunflower oil by the same 2. Each point of wax is a noticeable step.
• Softer: drop beeswax by 2 to 3 points into the liquid oil. For a tin, go to 9% wax as in the third table.
• Glossier: raise castor from 10% to 15%, taking it from the sunflower. Above 25% it turns tacky.
• Less greasy: replace 10 points of sunflower with squalane or fractionated coconut, both of which absorb dry and fast.
• Vegan: replace 16% beeswax with 10% candelilla, add the 6 points back as castor, and expect a drier feel unless you also swap shea for mango.
• Cheaper: cut jojoba to 5% and raise high-oleic sunflower to 44%, and use refined shea rather than unrefined. Expect shelf life to drop by a few months.

Troubleshooting

Where to go from here

Make the 25 g batch exactly as written and use it for a week before changing anything. That one batch tells you more about your wax and your preference for hardness than any amount of reading. When you do change something, change one number at a time and log it with the date, batch weight and ambient temperature.

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