Why Your Clean-Product Apps Disagree About Baby Wash (and How to Decide Anyway)

Published November 2, 2026 · Non-Toxic Baby & Screen-Free Play Directory

Yuka, EWG Skin Deep and Think Dirty often disagree about the same baby wash. Here's the published rule behind most of those disagreements, and how to scan a bottle and decide for yourself.

You scan the same bottle of baby wash in three apps and get three different verdicts: a green in one, an orange in another, a scary number in the third. It's easy to assume one app is wrong, or that the product is secretly bad. Most of the time, neither is true. Each app publishes its scoring rules, and the rules are different. Once you know the four rules behind most disagreements, a conflicting score stops being a crisis and becomes a clue.

The Three Apps in One Paragraph Each

Yuka rates cosmetics from 0 to 100 (higher is better). It sorts each ingredient into one of four risk levels (at risk, moderate, low, no risk), based on potential endocrine disruption, allergen, irritant, carcinogen and pollutant effects.^[1]

EWG Skin Deep gives a two-part rating: a 1–10 hazard score (lower is better) plus a data-availability rating of none, limited, fair, good or robust.^[4][5]

Think Dirty uses a 0–10 “Dirty Meter” (lower is better), built from evidence of carcinogenicity, developmental and reproductive toxicity, allergenicity and immunotoxicity, and environmental impact.^[6]

Rule 1: One Worst Ingredient vs. a Weighted Model

This is the biggest source of disagreement.

• Yuka: the highest-risk ingredient sets the range. If a high-risk (red) ingredient is present the product scores below 25/100; if the worst is moderate-risk (orange), it scores below 50/100. Other ingredients then subtract points (from 2 up to 12 each, depending on the type and level of risk).^[1][2]
• Think Dirty: by default, a product's rating is no lower than the highest rating of any single ingredient. It allows case-by-case exceptions, for example when 4–7-rated ingredients make up less than 5% of the formula, or when a product holds a credible certification.^[6]
• EWG: a product's hazard score is explicitly “not an average” of its ingredients. It uses a weight-of-evidence approach across 17 hazard categories with different weights, then scales scores from 1 to 10, with the top 5% most hazardous products and ingredients getting a 10.^[4][5]

What this means for baby wash: a wash can be built from mild cleansers and still contain one preservative or plant extract that one app rates as a concern. In Yuka or Think Dirty, that single ingredient can cap the whole product. In EWG, it is one input among many. When the apps disagree, tap into the product in each app and find the one ingredient that is dragging the score down. That's usually the whole story.

Rule 2: The Word “Fragrance”

• Think Dirty automatically rates any product that lists “fragrance” on its label at 7 or higher, because the individual chemicals in a fragrance blend aren't disclosed.^[6]
• EWG also treats undisclosed ingredients as a hazard: its methodology flags “ingredients not fully identified, such as fragrance and unidentified essential oils” as their own hazard category. But that category feeds into the weighted model rather than setting a fixed floor.^[5]
• Yuka evaluates fragrance like any other listed ingredient. We didn't find a fragrance-specific rule in its published cosmetics method.^[1][2]

What this means for baby wash: a scented wash can look fine in one app and land in Think Dirty's high range for that one word. For babies, many parents simply skip added fragrance. That's why every pick below is unscented or fragrance-free. Read the ingredient list, not the front label. Think Dirty's own methodology notes that products labeled “fragrance free” have been found to contain fragrance ingredients used to mask other odors.^[6]

Rule 3: A Low Score Isn't Always a Strong Score

EWG publishes how much research backs each rating. Its methodology says some ingredients rank low for hazards “only because little research” exists, and that the safest products score well on both measures: a low hazard score and a fair or better data-availability rating.^[4][5]

Think Dirty's lowest band, by contrast, groups together ingredients with “no concern or low concern identified, such as mild irritation, limited data, or use within regulatory limits.”^[6]

What this means for baby wash: washes with newer ingredients or unusual plant extracts can look great partly because there isn't much data on them. In EWG, check the data rating next to the hazard score. A low hazard score with “limited” data is a weak green, not a strong one.

Rule 4: None of the Apps Really Weighs Dose

• EWG: its hazard rating “does not account for exposure or individual susceptibility,” factors EWG says “will drive health risks, if any.”^[5]
• Yuka: its terms say the cosmetics method does not “take into account the quantity” of the ingredients.^[3]
• Think Dirty: argues that even small amounts matter because personal care products are used so often. The under-5% case-by-case exception in Rule 1 is the only place concentration comes in.^[6]

What this means for baby wash: an ingredient at a tiny fraction of a percent can move a score as much as one that makes up a large share of the bottle. The scores are a hazard screen, not a measure of how much your baby is actually exposed to. That's useful for spotting ingredients you'd rather avoid, but it's not a verdict on safety.

Your 5-Minute Referee Routine for Any Baby Wash

1. Read the label first. Look for “fragrance” or “parfum,” essential oils, plant extracts and preservatives (usually near the end of the list).
2. Scan it yourself in each app you use. Don't rely on screenshots or old reviews, since scores change.
3. Find the driver. In each app, open the ingredient breakdown and note the single ingredient with the worst rating.
4. Name the rule. Is the split caused by a worst-ingredient cap (Rule 1), the fragrance rule (Rule 2), thin data (Rule 3), or a trace ingredient that dose would change (Rule 4)?
5. Decide what matters to you. If your little one has sensitive or eczema-prone skin, allergen and irritant flags may matter more to you than long-term hazard flags. Your pediatrician can help with specific skin concerns.

Four Fragrance-Free Baby Washes to Scan Yourself

We chose these from their published ingredient lists: none lists “fragrance,” and each shows a different pattern you might see in the apps. We're not quoting any app's rating. Scan them yourself and use the notes to work out why the apps agree or disagree.

What the Apps Don't Tell You

• Your child's skin. A product with excellent scores can still irritate one particular baby, and the apps can't know that.
• Rinse-off and how you use it. How much you use, how often, and whether it stays on the skin aren't part of the published scores (see Rule 4).
• Label errors. App databases are built from labels and user submissions and can lag behind reformulations. Yuka's terms, for example, tell users to check the composition directly on the product packaging.^[3]

Frequently Asked Questions

Which app is right about my baby wash?

None of them is "right" in a single-number sense. They answer slightly different questions with different published rules. Use the app scores as a pointer to the one or two ingredients worth reading about, then decide based on what matters for your child, and ask your pediatrician about specific skin concerns.

Why does one app flag a baby wash that another app rates well?

Usually because of one of four published rules: a single worst ingredient setting the score (Yuka, Think Dirty) versus EWG's weighted model, Think Dirty's rule for anything listing "fragrance", EWG's data-availability rating, or the fact that none of the apps scales its score to how much of an ingredient is in the bottle.

Is "unscented" the same as "fragrance-free"?

Not always. Read the ingredient list instead of the front label and look for "fragrance", "parfum" or essential oils. Think Dirty's methodology also notes that products labeled fragrance-free have been found to contain fragrance ingredients used to mask other odors.

Why don't you show the app scores here?

The apps' terms don't allow their scores to be republished commercially, and a score copied today can change tomorrow. Scanning the bottle yourself takes seconds and gives you the current result plus the ingredient breakdown.

Sources

All methodology pages read 2026-11-02. Methods change; we re-check them periodically.

1. Yuka Help: How are cosmetic products evaluated?
2. Yuka Help: How are penalties calculated in the cosmetic product scores?
3. Yuka Terms & Conditions (USA), sections 3.2 and 6.3.1
4. EWG: Understanding Skin Deep ratings
5. EWG: Skin Deep methodology (About Skin Deep)
6. Think Dirty: Our Methodology

We are not affiliated with Yuka, EWG or Think Dirty, and app names are used only to describe their published methods. This article is general information, not medical advice. Patch-test new products, follow the manufacturer's directions, and ask your pediatrician about your child's skin.