Evidence & third-party tests
Every claim on this site carries a label saying where it comes from. We never blend these into one score, and we have no laboratory of our own — so nothing here is presented as "our test".
The five evidence tiers
Government agencies, university extension services, poison control. Safety rules come only from this tier.
Trade associations such as textile or stone institutes. Often expert, but they represent an industry — we say so.
Product-label directions and maker guidance, attributed to the maker. Commercial incentive noted.
Someone else's published hands-on test, linked. Explicitly not our test.
Forum threads and blog reports — linked, dated, and always labeled anecdotal. Never used for safety claims.
Third-party tests we cite
Babylist: A Babylist staffer stained a white baby footie with baby poop and pureed carrots, rinsed it in cold water, and laid it in direct sunlight for several hours, then repeated inside-out and machine washed.
Most of the stain faded within a few hours in the sun; after a second sunning and a wash cycle the garment came out completely clean. They suggest dish soap as a backup for cloudy days.
Newborn Poop Stains? All You Need Is Sunshine for This Magic Hack · 2026-03-25 · accessed 2026-08-18 · third-party test — not our test
Your Best Digs: In-house test of seven stain-remover products on white cotton shirts stained with coffee, wine, chocolate-hazelnut spread, and ink; stains were set for one week on the front and applied fresh on the back, then each product was used per label and washed identically.
OxiClean Versatile Stain Remover won overall, clearing all tested stains via soaking (ink faded substantially rather than vanishing). Tide To-Go pen was the portability pick and Shout Advanced Gel the gentlest on fabric.
The Best Laundry Stain Removers of 2026 · 2022-01-19 · accessed 2026-08-18 · third-party test — not our test
Cubby (via AOL): The author deliberately stained a white crib sheet with baby poop, ketchup, chocolate, coffee, and mud, then treated sections with five products: Amodex, Tide To Go pen, Puracy stain remover, Zote soap bar, and Miss Mouth's Messy Eater.
Puracy removed all stains except barely visible traces of mud and chocolate and was named the winner, with results improving with longer dwell time; Miss Mouth's performed similarly; Zote and Amodex ranked worst in this test.
I Tested 5 Stain Removers and There Was One Runaway Winner (I'm Never Without It Now) · 2025-09-17 · accessed 2026-08-18 · third-party test — not our test
Better Living: The author documented treating real grease stains (starting with fresh bacon grease on a linen shirt, plus older and heat-set oil stains on cotton and synthetics) with blue Dawn dish soap, baking soda, cornstarch/baby powder, and WD-40. Publication date not visible on page.
Dawn dish soap won: a fresh stain on cotton was completely gone after one treatment and a four-hour-old stain gone after two. Baking soda gave only 5-10% improvement, cornstarch mainly stopped spreading, and WD-40 gave roughly 60-70% improvement on set-in stains but was inconsistent and once spread the stain.
How to Get Grease Out of Clothes: What Actually Worked When I Ruined My Favorite Linen Shirt · accessed 2026-08-18 · third-party test — not our test
America's Test Kitchen: Stained white cotton T-shirts with 1/4 teaspoon of peanut oil, let it set 5 minutes, then compared 16 remedies: absorbents (cornstarch, cornmeal, baking powder, baking soda, talcum powder, salt), oddball treatments (hair spray, WD-40, cola, aloe vera, Cheez Whiz), and cleaners (Lestoil, white bar soap, shampoo, liquid dish soap). Publication date not visible on page.
Shampoo and liquid dish soap worked best on both fresh and overnight-set stains; none of the absorbents worked well and some spread the stain; cola and Cheez Whiz failed or added new stains; Lestoil underperformed and bar soap was only moderately effective.
Home Remedies for Grease Stains · accessed 2026-08-18 · third-party test — not our test
What we don't have (yet)
- Our own tests. A written protocol exists for future first-party testing; until real tests are run and photographed, no "editorially tested" claim appears anywhere.
- Outcome statistics. "Worked for N%" figures require real, moderated data collection. We publish none until that exists.