Last updated: August 18, 2026
The short version
Most “ingredients to avoid” lists are written to sell an alternative. They mix genuinely contested substances with harmless ones that happen to have long chemical names, and the result is a shopper who avoids the wrong things while feeling informed.
The honest position is narrower than the internet suggests. A small number of ingredients have real regulatory or evidential concerns worth acting on. A larger number are fine, and some of the “clean” replacements are more likely to irritate your skin than the things they replaced. Below is what actually holds up, and what does not.
Worth avoiding: the short, defensible list
Formaldehyde and formaldehyde-releasing preservatives
These appear as DMDM hydantoin, imidazolidinyl urea, diazolidinyl urea, quaternium-15, and a few others. They work by slowly releasing small amounts of formaldehyde, which is an effective preservative and also a recognized human carcinogen by inhalation and a well-documented contact allergen.
The case for avoiding them is not that a face cream will give you cancer. It is that they are a leading cause of contact dermatitis in patch testing, and equally effective alternatives exist. When there is a substitute with a better safety profile and no performance penalty, avoiding is the reasonable default.
Methylisothiazolinone and methylchloroisothiazolinone
Often written MI and MCI. These preservatives were adopted widely as parabens fell out of favour, and the result was a substantial rise in contact allergy across Europe and North America, significant enough that regulators restricted their use in leave-on products.
This is the clearest cautionary tale in the category: an ingredient swapped out on precautionary grounds was replaced by one with a demonstrably worse sensitization record. “Free from” is not the same as “safer”.
Oxybenzone in sunscreen
Oxybenzone absorbs systemically at measurable levels, has weak endocrine activity in laboratory studies, and is implicated in coral reef damage. Mineral filters, zinc oxide and titanium dioxide, work well and avoid all three questions.
The important caveat: the worst sunscreen choice is not using one. Ultraviolet exposure is a confirmed carcinogen with a far stronger evidence base than any filter concern. If a mineral sunscreen means you skip application, use the chemical one.
Fragrance, if your skin is reactive
“Fragrance” or “parfum” can legally represent dozens of undisclosed components, and fragrance mix is consistently among the top contact allergens in dermatology patch testing. For most people it is genuinely fine. For anyone with eczema, rosacea, or a history of reacting to products, it is the single most useful thing to eliminate first.
Note that this applies to natural fragrance too. Essential oils are fragrance. Limonene, linalool, citral, and geraniol are common allergens and occur naturally in citrus and lavender oils.
Genuinely contested, and reasonable to avoid if you prefer
Parabens. Methyl-, ethyl-, propyl- and butylparaben have weak estrogenic activity in laboratory assays, at levels far below the body’s own hormones. Regulatory bodies in the EU and US consider the common ones safe at permitted concentrations, and they are excellent preservatives with a low allergy rate. The MI story above is the reason to think twice before treating paraben avoidance as an obvious win.
Phthalates. Largely removed from personal care already, but can appear hidden inside “fragrance”. If phthalate avoidance matters to you, choosing fragrance-free is more effective than reading for the word.
Chemical UV filters generally. Beyond oxybenzone, most are well tolerated. Preference for mineral filters is reasonable; treating all chemical filters as dangerous is not supported.
Fine to ignore, despite the reputation
Sulfates (SLS, SLES). Detergents, not toxins. They can be drying and irritating at high concentration, particularly in facial cleansers, which is a comfort issue rather than a safety one. In shampoo, most people tolerate them well. There is no credible carcinogenicity evidence.
Mineral oil and petrolatum. Cosmetic grades are highly refined, inert, and among the least allergenic moisturizing agents available. Petrolatum is a dermatology staple for barrier repair precisely because it so rarely causes problems. The contamination concerns relate to unrefined industrial grades, not cosmetic ones.
Silicones. Large, inert molecules that sit on the skin surface. They do not penetrate, do not “suffocate” skin, and are non-comedogenic in most formulations. The objection to them is environmental persistence, which is a legitimate concern, but it is not a skin-safety one.
Preservatives as a category. This is the most consequential misunderstanding. Any product containing water needs preservation, and an unpreserved or under-preserved cream is a growth medium for bacteria and mould. Contaminated cosmetics cause real infections, particularly around the eyes. A preservative system is a safety feature.
Alcohol, sometimes. Fatty alcohols such as cetyl, stearyl, and cetearyl alcohol are emollients and the opposite of drying. Only the simple volatile alcohols, denatured alcohol or SD alcohol high in the ingredient list, are worth a second look on dry or sensitive skin.
Why the scary lists exist in the first place
It helps to understand where “ingredients to avoid” lists come from, because the pattern explains most of their errors.
Cosmetics are lightly regulated compared with drugs, and that regulatory gap created a genuine market opening for brands positioning themselves as safer. The commercial logic is straightforward: to sell an alternative, you need something for the customer to be avoiding. Long chemical names are excellent raw material, because unfamiliarity reads as risk.
The lists then propagate. A blog cites a brand’s marketing page, a second blog cites the first, and within a few cycles the claim has the texture of established fact without anyone having read a study. Substances that are structurally similar get merged, so a concern about one phthalate becomes a concern about all of them, and a hazard identified at industrial inhalation exposure becomes a warning about a leave-on cream at 0.3%.
This is not a conspiracy, and plenty of people repeating these lists believe them. But it does mean the source of a claim matters. A dermatology patch-test dataset showing which ingredients actually cause reactions in real patients is a different kind of evidence from a list of substances that sound alarming.
The tell is usually specificity. A credible concern names the substance, the concentration, the exposure route, and the population affected. A marketing concern names a category and leaves the rest to your imagination.
Hazard, exposure, and why the distinction matters
Almost every misleading skincare claim collapses the difference between hazard and risk, and the distinction is worth carrying around.
Hazard is the intrinsic capacity of a substance to cause harm. Risk is the probability of harm given actual exposure. A substance can be hazardous and pose negligible risk if the dose is tiny, the route is wrong, or it never reaches the tissue in question.
Formaldehyde is a useful worked example because it appears on both sides. It is a genuine carcinogen by inhalation at occupational concentrations, which is a real hazard. In a rinse-off shampoo containing a releaser at a fraction of a percent, the inhalation exposure is negligible. But the same ingredient is also a well-documented skin sensitizer at exactly the concentrations used in cosmetics, and that is a real risk to the skin. The cancer framing is the weak argument; the dermatitis one is strong.
The same logic runs the other way. Water is not hazardous, yet frequent washing damages the skin barrier and is one of the more common causes of hand dermatitis. Nothing about the ingredient list predicts that.
Applied practically: when you read that something is “linked to” a condition, the questions to ask are at what concentration, by what route, and in what population. If those three are missing, the claim has not told you anything actionable.
What actually causes reactions, according to the data
Dermatology maintains standardized patch-test series for diagnosing contact allergy, and the results are far more informative than any avoid-list, because they reflect what genuinely causes problems in people who present with them.
The consistent leaders across those datasets are metals such as nickel, fragrance mixes, preservatives including the isothiazolinones and formaldehyde releasers, and certain botanical extracts, with propolis, tea tree oil, and compositae plant family extracts appearing regularly.
Notably absent from the top of those lists are most of the ingredients that dominate popular avoid-lists. Silicones, mineral oil, petrolatum, and sulfates rarely feature as allergens, because they are either inert or irritants rather than sensitizers, and the distinction matters: an irritant causes a dose-dependent response in anyone exposed enough, while an allergen provokes an immune reaction in sensitized individuals at tiny doses.
The practical consequence is that a shopper optimizing against the popular list is often avoiding the inert ingredients while continuing to use the sensitizers. Someone with genuinely reactive skin who switches to a fragranced “natural” product containing essential oils has usually increased their exposure to the categories that actually cause trouble.
Building a routine around this
Knowing which ingredients matter is less useful than having a method, so here is one that works regardless of what the label says.
Change one thing at a time. Introducing three new products together makes any reaction impossible to attribute. Leave two weeks between additions when your skin is settled, longer if it is not.
Keep the leave-on products simplest. Exposure time drives most reactions. A cleanser is on your face for forty seconds; a moisturizer is on for eight hours. Spend your attention on the second one, and stop worrying much about the first.
Judge on outcome, not on the label. A product your skin likes is a good product for you even if the ingredient list would fail somebody’s purity test. The reverse is equally true, and a certified organic cream that leaves you red is not a success.
Keep the empty bottles until you are sure. If something goes wrong three weeks in, the ingredient lists of what you were using are the only diagnostic material you have. This sounds fussy and takes ten seconds.
Escalate properly. Persistent reactions are worth a dermatologist and a patch test rather than another round of substitution. Patch testing identifies the actual allergen from a standardized panel, which turns guesswork into a specific list of things to avoid for good. That is a genuinely different quality of answer from anything a product label can give you.
What “organic” does and does not tell you here
Skincare is where organic labeling is weakest, and it is worth being blunt about why.
The USDA National Organic Program is an agricultural standard. It applies cleanly to a single-ingredient product like cold-pressed jojoba oil, where the plant was farmed and the oil pressed. It applies poorly to an emulsion containing water, an emulsifier, a preservative system, and a thickener, none of which are farmed.
Some manufacturers voluntarily certify to the USDA standard, and those products follow the same tier rules as food, which we cover in organic label tiers. Many more use private standards such as COSMOS or Ecocert, which are designed for cosmetics and set their own thresholds for naturally derived content. Others simply print the word “organic” with nothing behind it, which is legal in a way it would not be on food.
So on skincare, the word alone carries less information than it does in the grocery aisle. What carries information is a named certifier, a short ingredient list, and a manufacturer willing to state concentrations.
It is also worth saying plainly: plant-derived does not mean gentle. Essential oils, citrus extracts, and botanical actives are among the more common irritants in skincare, and a formula stuffed with twenty botanical extracts has more allergen exposure than a plain one with four ingredients.
How to actually read a skincare label
Ingredients are listed in descending order by concentration, down to 1%. Below 1% they can appear in any order, which is why a marketing hero ingredient often sits near the end among the preservatives. If the thing on the front of the bottle is second-to-last on the back, it is present in a token amount.
The first five ingredients are most of the product. Read those and you know what you are buying. Everything after is refinement.
Short lists are easier to troubleshoot. Not inherently safer, but if you react to a product with six ingredients you can identify the culprit. With thirty-five you cannot.
“Free from” claims tell you what is absent, not what replaced it. Paraben-free tells you nothing about which preservative was used instead, and as the MI case showed, the replacement is sometimes worse.
Patch test anything new. A small amount on the inner forearm for a few days catches most reactions before they happen on your face. This is more useful than any ingredient list.
A sensible default position
If you want a rule that is short enough to use in a shop: prefer fragrance-free on anything that stays on your skin, avoid formaldehyde releasers and MI or MCI in leave-on products, use mineral sunscreen if you will actually wear it, and otherwise judge products on how your skin responds rather than on how the ingredient list reads.
Keep routines short. Most irritation people attribute to a single villain ingredient comes from using too many active products at once, particularly acids and retinoids layered without a gap. Simplifying almost always helps more than substituting.
Our researched picks in the categories where this comes up most are here: organic face moisturizer, organic sunscreen, organic body lotion, and organic body wash. Broader context is in our organic skincare guide, and how to read ingredient lists covers the mechanics in more depth.
Common questions about skincare ingredients
Are parabens actually dangerous?
The common ones are considered safe at permitted concentrations by regulators in both the EU and the US. They have weak estrogenic activity in laboratory assays at levels far below the body’s own hormones, and a low allergy rate compared with several replacements.
Is natural skincare gentler than synthetic?
Not reliably. Essential oils and botanical extracts are among the most common contact allergens in dermatology testing, while highly refined synthetics such as petrolatum are among the least. Source is a poor predictor of tolerance.
Should I avoid all preservatives?
No. Any water-containing product needs preservation, and contaminated cosmetics cause real infections. Choose which preservative system, rather than avoiding preservation.
Does “organic” mean anything on skincare?
Only when a certifier is named. USDA certification applies cleanly to simple plant oils but poorly to emulsions. Private cosmetic standards such as COSMOS or Ecocert are more common, and uncertified use of the word is legal.
Are sulfates bad for skin?
They can be drying at high concentration, which is a comfort issue rather than a safety one. There is no credible evidence of carcinogenicity, and most people tolerate them fine in rinse-off products.
What is the single most useful change for reactive skin?
Going fragrance-free on leave-on products, including natural fragrance and essential oils. Fragrance mix is consistently among the leading contact allergens in patch testing.
Organic Connection participates in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com. This article is general information, not medical advice. If you have a persistent skin condition, see a dermatologist.