We Read 1,846 Ingredients Across 54 K-Beauty Products. Here's What We Flagged.

At some point, everyone who orders Korean skincare asks the same quiet question: is Korean skincare safe? It usually shows up right after a friend forwards a headline about a sunscreen recall, or a video where someone waves an ingredient list around and says the word "toxins."

The marketing doesn't exactly calm things down. K-beauty packaging loves "EWG Green Grade," "derma-tested," "hypoallergenic," "clean" — the kind of claim printed next to a watercolor leaf. None of those phrases has a definition that would survive contact with a chemist.

So we did the thing this store exists to do. We read the ingredient lists. All of them. 54 products, 1,846 individual ingredient entries, each one checked against our glossary of more than 1,100 screened ingredients. Then we wrote the numbers down, including the ones that make us look less than perfectly pristine.

Here's how the screening works, what we flagged, and why 39 of our 54 products carry at least one flagged ingredient and we sell them anyway. With the flag showing.

Is Korean skincare safe? Start with who's checking

Korean cosmetics are regulated by the Ministry of Food and Drug Safety (MFDS), which maintains prohibited and restricted ingredient lists, much like the EU does. Korea also has a category the US doesn't: "functional cosmetics." Sunscreens, brightening products, and anti-wrinkle products have to submit efficacy and safety documentation before they can make those claims. That's a pre-market review step an ordinary American moisturizer never goes through.

The US system leans the other way. American cosmetics ran on self-substantiation for decades (the 2022 MoCRA law finally added facility registration and safety-substantiation requirements), while sunscreens sit in a separate lane as over-the-counter drugs. That drug lane is why the US filter list is stuck in the 1990s: when the FDA proposed updating its sunscreen rules in 2019, it proposed only zinc oxide and titanium dioxide as "generally recognized as safe and effective" and asked for more safety data on twelve older filters before re-affirming them [1]. Meanwhile, the newer filters Korean formulators actually prefer are approved in the EU and Korea but not yet approved for US OTC sunscreens. Not banned. Not condemned. Waiting on paperwork and studies. We wrote a whole piece on why Korean sunscreens feel different, and this regulatory gap is half of the story.

So here's our honest answer to "is Korean skincare safe": country of origin is a bad safety signal in both directions. Korean products aren't riskier than American ones, and a Korean flag on the box doesn't make an oxidized fragrance allergen any gentler. The useful unit of analysis is the ingredient. Which is where screening comes in.

What "Screened · Reviewed · Awarded" actually means

Our listing process has three gates, and a product has to clear the first two before it appears in the store at all.

Gate 1: we read the full INCI

Every ingredient, first to last, gets checked against our glossary, which currently holds more than 1,100 entries, and graded low, mid, or high risk. The grades are contextual. The same INCI name can grade differently in a rinse-off shampoo versus a leave-on cream, and a fragrance entry in a product aimed at compromised skin gets read harder than the same entry in a hand wash. One thing we want to be plain about: these grades are our screening's assessment, a structured way of reading a label. They are not a regulatory safety determination, and we don't pretend otherwise.

Gate 2: we check the claims against real reviews

An ingredient list tells you what's in the bottle, not whether the bottle does what the ad says. So the second gate pits marketing copy against what reviewers actually report: the "72-hour hydration" claim versus the person who says it pills under sunscreen. This gate also handles evidence triage. When a trending active's best studies were done by injection rather than in a cream, the topical claim gets downgraded no matter how glowing the reviews are; our PDRN article is the worked example of that rule.

Gate 3: awards, contested only among survivors

The category awards on our Awards page are only contested between products that already passed gates 1 and 2. An award here never rescues a formula the screening didn't like.

The numbers: 1,846 ingredients, graded

Across the 54 products currently in the catalog, this is the full distribution:

And the number we're proudest of publishing: 39 of 54 products carry at least one mid- or high-risk entry. That's 72% of our own catalog, flagged by our own screening. If a curation store tells you everything it sells is flawless, it's describing its reading effort, not its products.

Here's what actually got flagged most often at mid risk:

The high-risk column is a short and interesting list: fragrance graded high in 2 products, hydroxycitronellal in 2, and one product each for p-phenylenediamine (PPD), resorcinol, menadione, and Yellow No. 4. We went down a genuine rabbit hole on hydroxycitronellal, a floral fragrance component with enough patch-test history that EU rules require it to be named on labels rather than hidden inside "parfum" [2].

The flags we sell anyway (and print on the product page)

This is the part most stores skip, so let's not skip it.

Fragrance allergens: the fresh-cut apple problem

Linalool and limonene, the lavender-ish and citrus-ish terpenes in our mid-risk table, are barely allergenic when fresh. The problem is what air does to them. Think of a fresh-cut apple: fine at first, then browning within the hour. Fragrance terpenes do something similar, slowly reacting with oxygen to form hydroperoxides, and it's those oxidized forms that patch-test clinics keep finding reactions to. In one multicenter study, oxidized versions of common fragrance terpenes turned out to be frequent contact allergens in eczema patients [3], and later work found oxidized limonene and oxidized linalool reactions often show up together in the same people [4]. This is why the EU expanded its mandatory fragrance-allergen labeling from 24-ish named substances to more than 80 [2]. It's also why a half-used bottle from 2023 is not the same product it was when sealed. We flag these, we sell some of them, and the flag is on the page.

Salicylic acid: flagged because it works

Nine of our products contain salicylic acid, and we flag every one. Not because it's dangerous at cosmetic levels; the EU's scientific safety committee re-reviewed it and considers it safe at the concentrations cosmetics use [5]. We flag it because it's a genuinely active exfoliant, the kind of ingredient you should knowingly opt into rather than discover after your barrier complains. Same logic as putting the chili icon on a menu.

PPD: the one we flag loudest

Yes, we sell a hair-dye product containing p-phenylenediamine, the classic oxidative dye and one of dermatology's best-known contact allergens [6]. It's graded high risk, it says so in the open, and the product page tells you to patch test 48 hours before every single use. One more thing the marketing world blurs: "PPD-free" dyes often use the closely related PTD, which can cross-react in PPD-allergic people, so a PPD-free label is not an allergy-safe label. Nothing is. The full breakdown lives in our hair-dye shampoo article.

Menadione: the flag inside our best toner

Our most awkward flag: menadione, a synthetic vitamin K3 that's prohibited in EU cosmetics, sits inside a toner where the other 28 of 30 ingredients grade low risk. We could have quietly not mentioned it. Instead it's the first con on the product page, with a 48-hour patch test recommendation attached. That disclosure is the whole point of the brand: the screening isn't there to make our shelf look perfect, it's there so you know exactly what you're weighing.

What makes us reject a product

Rejections at gate 1 come from stacking, not single ingredients: a high-risk entry in a leave-on formula with no functional justification, or a fragrance-plus-essential-oil pileup in a product marketed at sensitive skin. Gate 2 rejections are simpler. If the reviews consistently contradict the headline claim, we don't list it, however pretty the ingredient list looks. A product has to pass both readings. Plenty don't.

Why we don't say "clean beauty"

You may have noticed we never use the phrase. "Clean" has no agreed definition; it's a vibe wearing a lab coat. In practice it usually means a free-from list chosen for marketing resonance, which produces strange outcomes: a "clean" product carrying seven unoxidized-for-now essential oils, while a boring, well-preserved formula gets side-eyed over a paraben with decades of safety data. Our framework is disclosure instead of purity theater. Grade everything, publish the flags, let you decide with your own skin history in hand.

Here's what those grades look like on three real product pages, from squeaky-clean to flagged-and-disclosed.

The practical version of all this: stop asking whether Korean skincare is safe as a category, and start asking what's in the specific bottle. Buy from anyone who shows you the full ingredient list and names their flags out loud. If a product lists fragrance, linalool, or limonene and your skin is reactive, remember the fresh-cut apple: keep the lid tight, don't stockpile, and toss the two-year-old half-empty bottle. Patch test anything with a disclosed high-risk flag, and patch test hair dye every single time, even a brand you've used before. And treat "clean," "non-toxic," and "chemical-free" as decoration. The ingredient list is the only part of the label that's legally required to tell the truth.