Argan oil or jojoba oil: which is better for the face?

In one lab at 120 C, jojoba resisted oxidation for 23.7 to 25.0 h and argan for 5.6 to 7.1 h. Argan has more vitamin E and 34% linoleic acid. Face balm verdict.

For a face balm that has to keep, jojoba is the better choice, and by a wide margin. In the one laboratory that tested both under the same conditions, three unrefined jojoba oils resisted forced oxidation at 120 C for 23.73 to 25.01 hours and three argan oils for 5.56 to 7.14. Argan is a conventional plant oil with about a third of its fatty acid as linoleic acid, the chain that starts rancidity; jojoba is about 97 percent wax esters with almost no polyunsaturated chains at all. Argan does carry more vitamin E than jojoba, which is the half of the competing advice that is true, but the tests show that its vitamin E does not close the gap. Use argan in a face product for its finish or its name, at a modest level and with an antioxidant, and let jojoba carry the liquid phase.

• Rancimat 120 C: jojoba 23.7-25.0 h, argan 5.6-7.1 h
• Rancimat 100 C: jojoba 39.8 h
• Argan linoleic: 34.1-34.3%
• Jojoba polyunsaturates: about 1%
• Tocopherols: argan 667 mg/kg, jojoba about 500
• Cosmetic argan oxidised 2.2 times faster than edible

The two oils, measured side by side

The pages that fill the first screen of results for this comparison are a private label manufacturer's blog, a beauty retailer's forum thread, a skincare brand's ingredient glossary and several copied pages on unrelated domains, one of them a spam subdomain on a university's web address. None of them gives a measured number for either oil. The table below assembles what the open literature does give, from the six sources listed at the foot of this page.

Two rows decide the choice for a face balm: the Rancimat rows, where jojoba is three to four times more stable in the only same laboratory comparison, and the polyunsaturate row, which explains why. Each oil on its own is covered in more depth on the argan oil and jojoba oil ingredient pages.

Which lasts longer: the Rancimat numbers

The Rancimat heats an oil in a stream of air and records the moment its volatile breakdown products start to accumulate. Metrohm, which makes the instrument, calls that time "the «induction time», which is a good indicator for the oxidation stability". A longer induction time means a more stable oil, at that temperature.

At 120 C, in one laboratory: three to four and a half times

The most useful single source is a table Metrohm published through AZoM in November 2020, listing induction periods for more than fifty natural cosmetic raw materials. Every oil in the rows below was run at 120 C in the same laboratory. The table does not state air flow or sample mass, so its figures should not be mixed with figures from other laboratories, but within the table the comparison is fair.

The ratio. The weakest unrefined jojoba (23.73 hours) lasted 3.3 times as long as the strongest argan (7.14). The strongest jojoba (25.01) lasted 4.5 times as long as the weakest argan (5.56). Even refined jojoba, which lost about a third of its stability against the unrefined ones, outlasted every argan oil in the table by more than twice.

Argan is not a fragile oil. The same table puts it ahead of both cold-pressed sweet almond oils above, by about two to nearly three times. Argan sits in the middle of the cupboard: well clear of almond, short of macadamia, and far short of jojoba. Where the rest of the liquid oils fall is set out in the oil comparison table.

The deodorised argan was the least stable of the three. At 5.56 hours it trailed the cold-pressed organic argan at 7.14. The table does not say why, and it covers one sample of each, so it does not prove that deodorising shortens argan's life. It is a reason to ask for an induction time on the lot you are buying rather than to assume that a refined, odourless argan will last longer.

The application note that leaves out the temperature

Metrohm's shorter application note AN-R-029, on the oxidation stability of cosmetic and pharmaceutical raw materials, lists "Argan oil, deodorized, org." at 5.56 hours, a mean of four determinations with a standard deviation of 0.10 hours. The note does not state the test temperature for that row; it says only that the method runs "at a constant temperature between 100–180 °C". On its own, then, the figure cannot be set against any other. The same description and the same 5.56 hours appear in the larger Metrohm table above, at 120 C, which is the only reason this page places it there. The note's use is its scatter: a relative standard deviation of 1.7 percent, small next to the gap between argan and jojoba.

At 100 C: why there is no argan figure in hours

Two open-access papers ran the Rancimat at the same conditions, 100 C and 20 litres of air per hour. Fagoaga and colleagues tested jojoba alongside olive, sunflower and rosehip, in quadruplicate, and reported induction periods: jojoba 39.80 hours, with a standard deviation of 3.46, ahead of olive at 27.50, sunflower at 10.39 and rosehip at 4.45. The authors say the ranking "correlated with PUFA levels", the polyunsaturated fatty acids.

Zaanoun and colleagues tested argan at 90, 100, 110, 120, 130 and 140 C, with 3 g samples in a Rancimat 743, in triplicate. But they did not report induction periods. Their Table 2 gives a reaction rate constant, k, for each oil at each temperature. For cosmetic argan at 100 C it is 36.66 ×10^-3 per hour.

The reciprocal of that number is about 27 hours, and it would be easy to set 27 hours against jojoba's 39.8. This page does not. The paper never defines k as the reciprocal of an induction period, never says how k was obtained from the instrument's output, and reports no induction period for either oil, so 27 hours is not its result. What the rate constants do support, with no conversion, is a comparison within the paper: roasted against unroasted argan.

Cosmetic argan is the less stable grade

Argan oil comes in two kinds, and the one sold for skin is the weaker of the two. Rabbaa's review describes the split: edible argan is pressed from roasted kernels, and "cosmetic argan oil is extracted from unroasted argan fruits and is primarily intended for skin and hair care". Zaanoun and colleagues made both from one harvest. They collected 600 kg of fruit at Ait Baha in Morocco, roasted 150 kg of the kernels at 110 C for 25 minutes, and pressed both batches mechanically.

At every temperature the cosmetic oil oxidised faster, by 2.2 to 2.7 times on the rate constants. The fatty acids cannot explain it: the paper found no statistically significant difference in saturated or unsaturated fatty acids between the two oils. What differed were the minor components. Roasting left the edible oil with more tocopherol, more than twice the beta-carotene and twenty times the phosphorus, a marker of phospholipids. The authors write that "a high content of β-Carotene, Phospholipid and Phosphorus would improve resistance to lipid oxidation", and that these contents explain the trends "to a certain extent", alongside "several other factors" including tocopherol. Their conclusion is direct: "edible argan oil can be stored much better than cosmetic oil".

That matters for this comparison because the cosmetic grade is the one a face balm uses. Roasted argan smells of toasted kernels, which is why the argan ingredient page advises against it in a balm. So the argan you would actually put on your face is the less stable version of an oil that already trails jojoba. Its Q10 for the cosmetic oil, the factor by which the rate rises per 10 C, was 2.16 across 90 to 140 C; room temperature was not tested.

Why jojoba is so much more stable

The difference is structural. Argan is a triglyceride oil: Rabbaa's review gives its saponifiable fraction as 99 percent triacylglycerols, three fatty acids on a glycerol backbone, with 34.10 percent linoleic acid in the analysis it tabulates. Zaanoun's cosmetic oil measured 34.30 percent polyunsaturated. Jojoba is not a triglyceride. The CIR describes it as "composed almost completely (97%) of wax esters of monounsaturated, straight-chain acids and alcohols", principally C20 and C22 chains, and the manufacturer figures reported by Fagoaga give 1 percent polyunsaturated.

Linoleic acid matters because each linoleic chain has a carbon sitting between two double bonds, a bis-allylic position, whose hydrogens are the easiest in the molecule for oxygen to strip off. That step starts autoxidation. Oleic acid, with a single double bond, has no such position. The full mechanism, and the measured ratios of how much faster linoleic oxidises than oleic, are in rancidity and oxidation. Roughly a third of argan's chains carry that weak point. Almost none of jojoba's do.

The CIR sums jojoba up in one sentence: it "is not easily oxidized and remains chemically unchanged for years". The same pattern, jojoba far ahead of a high linoleic plant oil, appears against sweet almond in jojoba against sweet almond oil, where the gap at 120 C in the same Metrohm table was 8 to 17 times.

Vitamin E: argan has more, and it does not decide the result

Here the competing advice gets one thing backwards and one thing right. A private label manufacturer's comparison page says that "Jojoba oil is particularly rich in vitamin E", with no source. A forum answer on a beauty retailer's site recommends argan instead because it "has vitamin e, fatty acids, antioxidants". On the measured numbers, the forum answer is closer.

The figures. The CIR states that jojoba seed oil "contains ~0.05% tocopherols", about 500 mg per kg. Zaanoun's cosmetic argan measured 667.04 mg per kg, by HPLC under AOCS method Ce 8-89, and the roasted oil 791.91. Rabbaa's review tabulates gamma-tocopherol alone at 715.42 mg per kg, with alpha at 42.23 and delta at 103.22. The review compiles these figures from earlier analyses rather than measuring them itself. Either way, argan carries more vitamin E than jojoba, and most of it is the gamma form. Jojoba is not "particularly rich" in it.

Why it does not settle stability. Argan has the larger antioxidant stock and still lost to jojoba by three to four times at 120 C. Fagoaga found the same thing from another direction: jojoba had the lowest total antioxidant capacity at 25 C of the four oils they tested and the longest induction time. The antioxidant does not need to be large when there is so little for oxygen to attack. In argan, the tocopherols are defending a third of the oil that is linoleic; in jojoba, they are defending almost nothing.

For a formulator this has one practical consequence. Argan's own tocopherols protect argan, not the rest of the batch. At 10 percent of a balm, 667 mg per kg of argan contributes about 67 mg per kg of tocopherol to the whole product. If the formula needs antioxidant protection, add it deliberately, as described in vitamin E and antioxidants, and consider rosemary extract alongside it.

Does jojoba "trick skin into producing less oil"?

No source we could find supports it. The claim on the pages in the results runs in two versions: that jojoba "helps trick the skin into producing less oil", and, in the forum thread, that because it mimics sebum "it can help control oil production on the skin". None of the pages that make it cites a study, and none names a measurement of sebum output.

The sources here do not test it either. The CIR's 2008 safety assessment, the most complete public review of jojoba, does not mention sebum at all. None of the composition or stability papers on this page measured sebum output after applying either oil.

What the claim borrows from is real, and narrower. Human sebum is about a quarter wax esters, a figure set out with its source on the tallow against shea butter page, and jojoba is nearly all wax esters. That is a match of chemical class. It is not evidence that skin senses the jojoba and switches its own production down, and no step between the two has been measured in the sources we hold. The same caution applies to the closest relative of this claim, the comparison with squalane, which is worked through in squalane against jojoba oil.

Why the claim persists

It sounds like a mechanism, it rests on one true fact about wax esters, and it answers the question oily skinned buyers actually have: will this make my face oilier? For whether any oil will block pores, see comedogenic ratings, which explains why those scores say little about a finished balm, and will a balm break me out.

Skin tolerance and allergy

Jojoba. The CIR reports that "in clinical tests, Simmondsia Chinensis (Jojoba) Seed Oil was neither a significant dermal irritant, nor a sensitizer", and the panel judged it safe "at concentrations up to 100% in body and hand creams". It also concluded that the jojoba ingredients "are not expected to easily penetrate skin".

Argan. None of the sources on this page reports a clinical irritation or sensitisation test for argan oil on human skin. Rabbaa's review is about antioxidant and anti-inflammatory effects in animal and cell models, and the skin finding it reports is from mice. So this page does not rank the two oils on tolerance, or repeat the anti-ageing claims made for argan on the competing pages.

Allergy. Argan comes from a kernel, and the site's argan page recommends treating it as a nut-derived oil for labelling purposes. What that means for a product sold to the public is set out in nut oil allergy. Jojoba is not a nut oil.

Which suits a face balm

A face balm is the product where stability matters most and is easiest to forget. It sits on skin you smell at close range, often from a jar kept in a warm bathroom. The site's face balm formula is built around that: its liquid phase is oleic-dominant and low in linoleic, with jojoba at 12 percent, squalane at 25 and 0.1 percent mixed tocopherols for the sunflower fraction.

The arithmetic of a blend. In a blend, the least stable oil sets the pace, so a balm's liquid phase is only as durable as its weakest major component. At 5 percent of the batch, argan's 34 percent linoleic acid puts about 1.7 percent linoleic into the whole formula; at 20 percent it puts in about 6.8. The oil blend calculator does this sum for any combination and gives a weighted stability index.

Where argan earns its place. The argan ingredient page suggests 10 to 20 percent in a face balm. The numbers on this page argue for the low end of that, 5 to 10 percent, unless the product will be used up within months. Above that, you are trading measurable shelf life for a sensory difference that a balm's wax and butter may bury. Whether testers can actually tell an argan version from one without it is a question to settle with a blind triangle test, as in sensory testing, before you pay for it.

Which argan to buy. Unroasted cosmetic grade, because roasted oil smells of food, while accepting that it is the less stable grade. Ask the supplier for the lot's peroxide value and, if they have one, a Rancimat induction time and the temperature it was run at. A supplier who quotes an induction time without a temperature, as Metrohm's own application note does, has not given you a number you can compare. More questions to put to a supplier are in sourcing ingredients, and the smell of an argan lot that has already turned is described in rancid balm.

What these numbers do not settle

Stability. The only same laboratory comparison is Metrohm's 120 C table, which does not state air flow, sample mass or number of runs for each row, and lists one sample per description. The 100 C comparison cannot be completed in hours, because the argan paper reports rate constants without saying how they relate to induction time. The roasted against unroasted result comes from one harvest in one region. All of it points the same way, and none of it predicts a shelf life in months.

Composition. Argan's rows come from a review of earlier analyses and from one harvest; jojoba's polyunsaturate figure is the manufacturer's. The certificate of analysis for the lot you buy outranks every figure here.

Skin effects. No source here tested either oil on sebum output, pore blocking, hydration or ageing in people, and only jojoba has published clinical irritation data in the sources we hold. This page compares the oils as materials. What either does on a particular face is not something these sources measured.

Reviewed and updated 13 October 2026. Spotted an error? Tell us and we will fix and log it.