Mineral Oil Is an Occlusive, Not a Humectant. The 98% Figure Has No Source We Could Read

In March 2025 this site published an article about luxury skincare that said a jar of La Mer contains “common humectants like glycerin and mineral oil”. Glycerin is a humectant. Mineral oil is an occlusive, which is close to the opposite. On 9 September 2026 we corrected it, and the article now reads “a humectant like glycerin and an occlusive like mineral oil”.

A small fix. But the two words describe opposite ways of keeping skin from drying out, and swapping them leads to wrong conclusions about what a product can do, so we went back to the sources to set the difference out properly. On the way we found that the most repeated number in this corner of skincare, that petrolatum cuts water loss through the skin by 98 or 99 percent, leads through a chain of reviews quoting each other to papers that either do not contain it or that we could not read.

Two opposite ways to stop skin drying out

An occlusive sits on or in the top layer of skin and slows the rate at which water leaves. It adds none. A humectant attracts water and holds it. Mineral oil, petrolatum and lanolin are usually filed as occlusives; glycerin, urea and hyaluronic acid as humectants. One slows the exit, the other holds on to what is there.

United States regulation treats both as legitimate. Under the FDA’s over-the-counter skin protectant monograph, a product can be sold as a drug for protecting chapped or cracked skin if its active ingredient is on a fixed list, within a fixed range of strengths. Occlusives and a humectant share that list, with very different minimums.

Ingredient	Usually filed as	Permitted strength as a skin protectant
Petrolatum	Occlusive	30 to 100%
Mineral oil	Occlusive	50 to 100%
Dimethicone	Occlusive (disputed, see below)	1 to 30%
Glycerin	Humectant	20 to 45%

That strength goes on the label. A skin protectant carries a Drug Facts panel with its active ingredient as a percentage, which is more than any cosmetic has to tell you. Aquaphor Healing Ointment declares “Petrolatum 41%”; Vaseline Healing Jelly declares white petrolatum at 100%. A cosmetic label, by contrast, only gives you order, and our guide to ingredient lists explains why. The La Mer cream that started all this lists mineral oil and petrolatum near the top; our reading of its three ingredient lists covers what that does and does not tell you.

The 98% figure, and the papers it points to

Read about petrolatum for long enough and you will meet a number. It usually arrives with a citation. We followed the citations through five reviews and a clinical reference, and they end at three papers. The first two are cited directly for the figure.

The first is a 2017 review in Clinical Medicine and Research, which says petrolatum at “a minimum concentration of 5%, can reduce trans-epidermal water loss by more than 98%”. Its reference for that sentence is a 1992 paper by Ghadially, Halkier-Sorensen and Elias in the Journal of the American Academy of Dermatology.

We could not get that paper’s full text; the publisher’s site refuses us. Its abstract contains no percentage of any kind. Nor is it a study of how much petrolatum reduces water loss. It asked where petrolatum goes in the skin and what it does to barrier repair, and found that petrolatum “accelerated, rather than impeded, barrier recovery”. Its conclusion even cuts against the usual picture of a seal on the surface: petrolatum “neither forms nor acts like an epicutaneous impermeable membrane”, and instead spreads through the spaces between cells in the outer layer.

The second is StatPearls, a clinical reference on the National Library of Medicine’s bookshelf, which calls petrolatum the most effective occlusive, “reducing water loss through the epidermis by nearly 99%”. Its reference for that sentence is a 2017 study of how a drug called ufenamate passes through pig skin from four oily bases. The abstract says nothing about water loss.

The same sentence turns up elsewhere with the same numbers. A 2005 article in Skin Therapy Letter says “Petroleum jelly, in a minimum concentration of 5%, reduces TEWL by more than 98%”, and its reference list includes the Ghadially paper and a 2001 article in the same journal. That 2001 article says petrolatum “in a minimum concentration of 5% is the most effective occlusive”, and gives no percentage for water loss at all. So the 5% is older than the 98%.

A 2016 review in the Indian Journal of Dermatology repeats it once more, “Petrolatum in a minimum concentration of 5% reduces TEWL by more than 98%”, and cites two sources: the 2001 article and a 1971 study by Spruit in Dermatologica. That is the end of the trail, and the only measurement on it. Its abstract says “The specific resistances (SR) of various substances (a.o. vaseline, wool-fat, emulsions, paraffins, vegetable oils, simple esters) to the passage of water were measured.” It also reports measuring how far some of them reduced water loss from skin, but it gives no figures, and the paper is a four-page PDF behind a paywall that we did not buy.

So here is where it stands. The reviews repeat one another. The paper they cite most often has no percentage in its abstract; we read the abstract, not the paper, so we cannot rule out a figure in its results. The one measurement the trail reaches may contain the number, but we could not read it, and none of the reviews quotes its figures. Until someone shows the 98% in a study, it is a number without a source anyone has put on the page. Our La Mer article declined to print it for the same reason.

What has actually been measured

The best measurement we could reach comes from Patzelt and colleagues at the Charité in Berlin, published in 2012. Its abstract describes six healthy volunteers, four plant oils and paraffin oil (another name for mineral oil) applied to the skin, with petrolatum as the comparison. It found “The most effective occlusion was found for petrolatum.” and described the oils as producing “a semi-occlusion of the skin surface”.

The paper itself is paywalled, but its numbers are reproduced in a table in a 2012 review of mineral oil by Rawlings and Lombard, which we read in full. Water loss was measured before application and 30 minutes after. Our arithmetic from that table:

Petrolatum roughly halved water loss in that test. Mineral oil trimmed it by about a tenth. The same review quotes other figures pointing the same way, with “mineral oil reducing baseline transepidermal water loss (TEWL) by 16% (for comparison, petroleum jelly reduced TEWL by 43%)”, though it does not make clear which study those come from. Nothing we found reaches 98%.

Three things to know about that review. First, who paid for it: “This review has been fully funded by Union Swiss.” The second author works at Union Swiss and the first consults for it. As we found with azelaic acid, a funded review is not a wrong one, and here the funding is declared and the table reproduces someone else’s data. Second, these are single applications measured within half an hour on a handful of people, not a trial of daily use. Third, set the review’s summary against its own numbers: its comparison table rates mineral oil’s occlusivity “High (because of alignment of straight alkyl chains)”, while the figures it prints for mineral oil on skin run from about 10 to 25 percent.

Dimethicone is the surprise. It is routinely listed with the occlusives, StatPearls included. A 2014 study of 26 women at the Vrije Universiteit Brussel compared three common silicones with petrolatum on forearm skin and reported that “Both petrolatum and silicones significantly decreased TEWL 15 min after application, but the measurements for the silicones were not significantly different from the untreated control values”. The paper’s title says it plainly: “Silicones as nonocclusive topical agents.”

Mineral oil is not petrolatum

The two come from the same place and get treated as one ingredient in most writing, including the sentence we corrected. They behave differently. Petrolatum is a semi-solid mixture; mineral oil is a liquid. In the Patzelt data petrolatum was about five times as effective at slowing water loss 30 minutes in. The FDA monograph reflects the gap in its own way: petrolatum counts as a skin protectant from 30%, mineral oil on its own only from 50%.

So the corrected sentence in our older article is right, and if anything generous. Mineral oil is an occlusive. On the measurements we could find, a mild one.

Does mineral oil clog pores?

The worry comes from animal testing. Raw cosmetic ingredients were screened on the ears of rabbits for their tendency to form plugged follicles (comedones), and the results became lists of ingredients for acne-prone people to avoid, mineral oil among them. The founding paper, Kligman and Mills’s 1972 “Acne cosmetica”, has no published abstract and we could not read it. The numeric scores for mineral oil quoted online disagree with each other so badly that we are not repeating any of them.

What we can read is what the same two researchers said ten years on. In 1982 Mills and Kligman published a human version of the test, applying substances under occlusion to the upper backs of young men with large follicles for a month. Their abstract says “The rabbit model is more sensitive than the human.” and “Substances that are weakly comedogenic in the rabbit are probably safe for human use with the possible exception of acne-prone persons.” The same year, a review by Frank was harsher, pointing to “the absence of correlation with experience in the human”.

The case in mineral oil’s favor is thinner than it usually sounds. These are the two papers usually cited for it.

• DiNardo, 2005. A two-page communication titled “Is mineral oil comedogenic?” in the Journal of Cosmetic Dermatology. It has no abstract and we could not read it. The publisher’s press release summarizes it as saying “while industrial grade mineral oil may be comedogenic, cosmetic grade mineral oil is not”. The author’s listed affiliation is a private research firm, Pharma Cosmetix Research.
• Draelos and DiNardo, 2006. A study in the Journal of the American Academy of Dermatology that used a modified version of the human back test on six people, with patches of finished cosmetic products, not mineral oil on its own. Its conclusion: “Finished products using comedogenic ingredients are not necessarily comedogenic.”

You will read that cosmetic-grade mineral oil scores zero in human testing. The fullest version of that claim we found is in the Union Swiss review discussed above. It says Mills and Kligman ran mineral oil through both their rabbit and human models, “in which mineral oil was reported not to be comedogenic”, that Fulton’s 1989 list rated it mildly comedogenic but far less so than other ingredients, and that DiNardo tested formulations with up to 30% mineral oil in people and found no comedogenic potential. The 1982 abstract does not name the substances it tested, and we could not read that paper or DiNardo’s, so for those results we are relying on a funded review’s account. What the readable evidence supports is narrower: the test that put mineral oil on the list is known to over-predict, and nobody we found has shown that refined mineral oil clogs human pores. That is not the same as proof that it never does, particularly for acne-prone skin, which is the exception Mills and Kligman made themselves.

Industrial mineral oil and cosmetic mineral oil

The other fear about mineral oil is cancer, and it rests on a real distinction. The International Agency for Research on Cancer classifies “Mineral oils, untreated or mildly treated” in Group 1, carcinogenic to humans. “Mineral oils, highly-refined” are in Group 3, not classifiable. These are different materials that share a name.

Germany’s Federal Institute for Risk Assessment reviewed highly refined mineral oils in cosmetics in a 2018 opinion and concluded that “no health risks are to be expected for consumers who apply cosmetic products to their skin”. Lip products, which get swallowed in small amounts, get a separate discussion. There the opinion expects no health effects either, provided manufacturers follow an industry recommendation to use only mineral oil grades that have an acceptable daily intake.

Do humectants pull water out of your skin in dry air?

This is the other half of the distinction, and the claim most often attached to it. StatPearls puts it this way: humectants can hydrate skin from the air when humidity is high, but “more commonly, however, they draw water from the deeper epidermis and dermis”. That sentence carries no citation in the text we read.

We looked for a measurement on human skin: a humectant applied in a controlled dry room, with water loss or skin hydration measured against untreated skin. We did not find one. The nearest tests we found point in different directions, and none of them settles it.

• On forearms under drying cups, glycerin raised water loss. In a 1959 paper, researchers at the Warner-Lambert Research Institute strapped small cups of silica gel, a drying agent, to volunteers’ forearms for two hours. Anhydrous glycerin raised the water the cups collected by 43%, while petrolatum cut it by 48%. The authors concluded that “many humectants and hand lotions tend to increase the rate at which the skin dries out”. The numbers of determinations were small, three to eight per ingredient.
• At ordinary room humidity, glycerin lowered it. In a 1988 study from Beecham Products in England, 5% and 15% glycerol in water lowered water loss on forearms compared with plain water, in rooms kept no more humid than 60%. The authors wrote that this “contradicts that published data (19,20) which was obtained either in vitro, where normal TEWL is lacking, or under conditions of abnormally low or zero external relative humidity”, one of those references being the 1959 paper.
• Glycerin may not work mainly by holding water. A 2014 study from Lund University on samples of the skin’s outer layer, which a later paper from the same group describes as pig skin, noted that natural moisturizing factors are “commonly claimed to increase the water content in the SC and thereby protect the skin from dryness. In this work we challenge this proposed mechanism”. Urea and glycerol kept the skin’s outer layer more mobile at moderate humidity, and “This effect cannot be attributed to increased SC water content.”
• The same finding in skin lipids. A 2024 study of isolated human skin lipids concluded that “glycerol enhances lipid cohesion rather than increasing water uptake, which is typically the expected function of humectants”.
• In mice, a humectant protected against dry air. A 1998 study from the Shiseido Research Center kept mice in very dry air, then applied a plastic film, petrolatum or a humectant. “All three forms of treatment prevented the epidermal hyperplasia and dermal mast cell hypertrophy and degranulation induced by exposure to low humidity.” The humectant helped rather than harmed, in that model.

So glycerin under very dry conditions has been measured raising water loss from living skin, in one small test from the 1950s, and a later study at normal humidity found the reverse. The popular version, that humectants pull water up from the deeper layers of your skin, goes further than any of these tests: none of them measured where the water came from. We come back to it, with hyaluronic acid, in a later article.

Why the distinction still matters

If the two worked the same way, it would not matter which one a sentence named. They do not. An occlusive does nothing to attract water; a humectant does little to stop it leaving. That is the textbook reason moisturizers so often carry both, and the labels bear it out. Aquaphor pairs its petrolatum with inactive ingredients that include “Mineral Oil, Ceresin, Lanolin Alcohol, Panthenol, Glycerin, Bisabolol”. CeraVe Healing Ointment declares “Petrolatum 49.5%” with mineral oil and glycerin among its inactives.

It also changes how to read a claim. Johnson’s Baby Oil breaks its formula down on its own page as about 99% mineral oil and under 1% fragrance, and the same page says it “locks in up to 10x more moisture on wet skin than an ordinary lotion can on dry skin”. The page gives no study for that figure. Mechanically the idea makes sense: an occlusive has no water of its own to hold, so putting it on damp skin gives it something to hold in. The number is another matter.

So what is mineral oil good for

• Mineral oil: occlusive, and a mild one.
• Petrolatum: the stronger occlusive, about half in one small test.
• The 98% figure: no source we could read.
• Pores: a rabbit test that over-predicts.
• Humectants in dry air: one small old test, glycerin only.

Where to find these

The links below are affiliate links, and the disclosure above the article applies to all three. The first two are sold as over-the-counter skin protectants with the strength on the label; the third is a cosmetic. None of them has been tested by us, and the measurements above are of the ingredients, not of these products.

• Vaseline Original Healing Jelly (paid link)
• Aquaphor Healing Ointment (paid link)
• Johnson’s Baby Oil (paid link)

References

• United States Food and Drug Administration. OTC Monograph M016: Skin Protectant Drug Products for Over-the-Counter Human Use, posted 24 September 2021. accessdata.fda.gov. Read in full.
• Drug Facts labels on DailyMed for Vaseline Healing Jelly, Aquaphor Healing Ointment and CeraVe Healing Ointment. Read in full.
• Purnamawati S, Indrastuti N, Danarti R, Saefudin T. The Role of Moisturizers in Addressing Various Kinds of Dermatitis: A Review. Clinical Medicine and Research 2017;15(3-4):75-87. PMC. Read in full.
• Ghadially R, Halkier-Sorensen L, Elias PM. Effects of petrolatum on stratum corneum structure and function. Journal of the American Academy of Dermatology 1992;26(3 Pt 2):387-396. Europe PMC. Abstract read; full text not read.
• Harwood A, Nassereddin A, Krishnamurthy K. Moisturizers. StatPearls, last updated 12 February 2024. NCBI Bookshelf. Read in full.
• Kraft JN, Lynde CW. Moisturizers: what they are and a practical approach to product selection. Skin Therapy Letter 2005;10(5):1-8. skintherapyletter.com. Read in full.
• Lynde CW. Moisturizers: what they are and how they work. Skin Therapy Letter 2001;6(13):3-5. skintherapyletter.com. Read in full.
• Sethi A, Kaur T, Malhotra SK, Gambhir ML. Moisturizers: The Slippery Road. Indian Journal of Dermatology 2016;61(3):279-287. PMC. Read in full.
• Spruit D. The interference of some substances with the water vapour loss of human skin. Dermatologica 1971;142(2):89-92. Karger. Abstract read; full text not read.
• Iino H, Fujii M, Fujino M, et al. Influence of Characteristics of Oily Vehicle on Skin Penetration of Ufenamate. Biological and Pharmaceutical Bulletin 2017;40(2):220-226. Europe PMC. Abstract read.
• Patzelt A, Lademann J, Richter H, et al. In vivo investigations on the penetration of various oils and their influence on the skin barrier. Skin Research and Technology 2012;18(3):364-369. Europe PMC. Abstract read; its table read as reproduced by Rawlings and Lombard.
• Rawlings AV, Lombard KJ. A review on the extensive skin benefits of mineral oil. International Journal of Cosmetic Science 2012;34(6):511-518. Wiley. Read in full. Funded by Union Swiss.
• De Paepe K, Sieg A, Le Meur M, Rogiers V. Silicones as nonocclusive topical agents. Skin Pharmacology and Physiology 2014;27(3):164-171. Europe PMC. Abstract read.
• Mills OH Jr, Kligman AM. A human model for assessing comedogenic substances. Archives of Dermatology 1982;118(11):903-905. Europe PMC. Abstract read.
• Frank SB. Is the rabbit ear model, in its present state, prophetic of acnegenicity? Journal of the American Academy of Dermatology 1982;6(3):373-377. Europe PMC. Abstract read.
• DiNardo JC. Is mineral oil comedogenic? Journal of Cosmetic Dermatology 2005;4(1):2-3. Wiley. Not read; no abstract is published. Publisher’s press release, ScienceDaily, read.
• Draelos ZD, DiNardo JC. A re-evaluation of the comedogenicity concept. Journal of the American Academy of Dermatology 2006;54(3):507-512. Europe PMC. Abstract read.
• International Agency for Research on Cancer. Agents Classified by the IARC Monographs, Volumes 1 to 123. iarc.who.int. Read.
• German Federal Institute for Risk Assessment (BfR). Highly refined mineral oils in cosmetics: health risks are not to be expected according to current knowledge. Opinion No 008/2018. bfr.bund.de. Read in full.
• Björklund S, Andersson JM, Pham QD, et al. Stratum corneum molecular mobility in the presence of natural moisturizers. Soft Matter 2014;10(25):4535-4546. Europe PMC. Abstract read.
• Sagrafena I, Morin M, Paraskevopoulos G, et al. Structure and function of skin barrier lipids: Effects of hydration and natural moisturizers in vitro. Biophysical Journal 2024. Europe PMC. Read in full.
• Denda M, Sato J, Tsuchiya T, Elias PM, Feingold KR. Low humidity stimulates epidermal DNA synthesis and amplifies the hyperproliferative response to barrier disruption. Journal of Investigative Dermatology 1998;111(5):873-878. Europe PMC. Abstract read.
• Powers DH, Fox C. The Effect of Cosmetic Emulsions on the Stratum Corneum. Journal of the Society of Cosmetic Chemists 1959;10:109-116. SCC library. Read in full.
• Batt MD, Davis WB, Fairhurst E, Gerrard WA, Ridge BD. Changes in the physical properties of the stratum corneum following treatment with glycerol. Journal of the Society of Cosmetic Chemists 1988;39:367-381. SCC library. Read in full.
• Johnson’s Baby Oil product page, read 16 September 2026. johnsonsbaby.com.