Restricted Substances Testing for Knitwear: What to Test and When
Quick answer
An RSL test checks that a finished garment contains none of the chemicals your market bans above set limits. For knitwear the realistic scope is azo dyes and aromatic amines in dyed yarn, formaldehyde in resin and anti-pilling finishes, pH, nickel release from metal trims, PFAS in any water-repellent finish, and lead, cadmium and phthalates in prints and buttons. Budget roughly USD 120–400 per composite package at a third-party lab, run it on a pre-production sample rather than on a lab dip, and re-test whenever the yarn dye lot, trim supplier or finishing recipe changes.
Most buyers meet restricted substance testing the same way: a retailer asks for a test report, the factory sends a PDF, nobody reads it properly, and two years later a shipment gets pulled. We see this a lot. The test itself is not complicated — what trips people up is scope and timing. A report that covers the wrong colourway, or was run on a sample made from a different dye lot, is a piece of paper that will not protect you.
What an RSL actually is — and what it is not
An RSL (restricted substance list) is a buyer's or retailer's list of chemicals that may not appear above a stated limit in the finished product. It is not a certificate and it is not a factory audit. It is a chemical test on a physical garment, run by a lab, producing a report with numbers and detection limits.
Two related terms get mixed up constantly:
- MRSL (manufacturing restricted substance list) — covers what may be used in the factory process, including chemicals that never stay on the garment. ZDHC's MRSL is the common one. It is about inputs and wastewater, not about the sweater in the box.
- RSL — covers what is actually present on the finished garment. This is what a physical lab test measures.
Certifications such as Oeko-Tex Standard 100 sit somewhere in between: they involve testing, but they certify against the scheme's own limits rather than your retailer's list. Our notes on knitwear certifications go into what each one does and does not cover. If your customer has their own RSL, their own limits apply — a certificate alone is not proof of conformance.
The substances knitwear buyers should actually test for
A blanket "test everything" request gets expensive fast. In practice, these are the ones that matter for knitted garments:
- Azo dyes / aromatic amines. Certain azo colorants can cleave into banned amines. The risk concentrates in strong reds, oranges, browns and blacks from uncontrolled dyestuff sources. This is the single most-tested item on knitted apparel.
- Formaldehyde. Present in resin finishes — including the easy-care and anti-pilling finishes sometimes applied to cotton and viscose knits, and in some pigment prints. Limits for children's wear are far tighter than for adults.
- pH. Skin-safety related. Dark and heavily dyed shades drift alkaline if neutralisation after dyeing is rushed. It is a cheap test and one of the most common soft failures.
- Nickel release. Applies to metal trims in prolonged skin contact — zips, buttons, eyelets, metal badge plates. This is a trim problem, not a yarn problem, and it is where most knitwear failures hide. See trim card approval for how to get trim documentation early.
- PFAS. Only relevant if the garment has a water-repellent or stain-release finish. Legislation in several US states and in the EU is tightening quickly, so if you ordered a "water resistant" knit, test for it rather than assuming.
- Lead, cadmium and phthalates. Mainly prints, plastisol transfers, coated buttons and rubber or PVC trims.
- APEO / NPEO residues. Surfactants from scouring and washing. Usually fine with modern auxiliaries, worth including if your market is the EU.
Colour fastness, shrinkage and pilling are physical tests, not RSL tests, but they are often booked on the same submission. Our notes on lab testing standards cover that side separately.
When to test — the timing most buyers get wrong
The instinct is to test the first sample. That is usually too early and always too narrow, because the first sample is often made in stock yarn and stock trims that are not what bulk will use.
- Do not test a lab dip. A lab dip is yarn or a swatch. It tells you about the dye, not the finished garment, and it cannot cover trims.
- Test the pre-production sample. This is the first sample made with the actual bulk yarn, actual dye lot and actual trims. It is the earliest point at which a report means anything.
- Re-test when inputs change. New dye lot, new yarn spinner, new trim supplier, or a change from piece dyeing to yarn dyeing — all invalidate the old report. Our piece on yarn-dyed versus piece-dyed knitwear explains why the two routes carry different chemical loads.
- Allow a re-test window. Lab turnaround is typically 4–7 working days, plus a day or two for submission and courier. Build that into the calendar before you promise a delivery date, not after.
How many tests is enough?
Labs allow composite testing: several colourways of the same material and construction are combined into one sample and tested as a group. If four colours come off the same yarn base with the same dyestuff family, one composite covers all four at roughly the price of one to one-and-a-half tests. If one of those colours is a deep black with a separate dye recipe, test it separately.
Rules of thumb that hold up in practice: composite by yarn and dyestuff family, not by your internal colour name; always test the darkest and most saturated shade separately; and always test metal trims as a separate line item, because trims fail independently of the knit.
How to read a test report
A report is only useful if three things line up:
- The sample description matches bulk. Fibre composition, colour, style number. If the report says "100% cotton" and your garment is a 95/5 cotton-cashmere, it is not your report.
- The limits are your customer's limits. Labs report against whichever standard you ask for — GB, EN, CPSIA, or a retailer's list. "Pass" against the wrong list is not a pass.
- The detection limit is stated. "Not detected" is meaningless without the method detection limit. A result below the limit with a stated method is fine; a bare "ND" with no method is a red flag.
Also check the date. Most retailers treat apparel test reports as current for about 12 months; beyond that they want a fresh one, especially if anything in the supply chain moved.
What actually fails, in our experience
Not the yarn, most of the time. The pattern we see at our partner factories in Dongguan:
- Metal trims. A zip pull or shank button sourced at the last minute with no test documentation. Fix: freeze trims at the pre-production stage and collect the trim supplier's report then, not at packing.
- pH drift on dark shades. Rushed neutralisation. Fix: it is a process control issue and the fix costs nothing if caught before bulk washing.
- Formaldehyde from an added finish. A buyer asks for "easy care" or "anti-pilling" without knowing it means a resin finish. Fix: state the performance requirement, and ask what chemistry delivers it.
- Prints and plastisol transfers. Added late, tested never. Fix: if it is printed, it needs its own line on the test submission.
Costs and what we do about them
Realistic figures for a standard knitwear package at an international third-party lab (SGS, Intertek, TÜV, Bureau Veritas): USD 120–250 for a core package of azo, formaldehyde, pH and nickel on a composite, rising to USD 250–400 once PFAS, phthalates and heavy metals are added. Rush service adds 50–100%. Per-garment, spread over a 600-piece order, that is cents — which is why testing late and failing is so much more expensive than testing early.
At YZR Sweaters we work with certified partner factories (BSCI, SEDEX, Oeko-Tex) across Dongguan, and we ask for the chemical question at quotation stage rather than at packing: which yarn, which dyestuff source, which trims, and whether any performance finish is requested. That conversation determines the test scope, and knowing it early means the pre-production sample is submitted for testing while there is still time to react. If your market has its own RSL, send it with the enquiry and we will quote the scope against it.
Planning an order that needs chemical compliance documentation? Tell us your market, fibre and trim plan, and we'll tell you what the realistic test scope is — before you commit to bulk.
Get a free quote Browse our knitwear rangeFrequently asked questions
Do I need a separate test report for every colourway?
Usually not — labs allow composite testing, combining several colourways of the same yarn base and dyestuff family into one submission at roughly the cost of one to one-and-a-half tests. Always test the darkest or most saturated shade separately, and always test metal trims as a separate line, because trims fail independently of the knit.
Is an Oeko-Tex certificate the same as an RSL test?
No. Oeko-Tex Standard 100 certifies against that scheme's own limits, which may be looser or tighter than your retailer's list, and a certificate covers what was submitted at the time of certification. If your customer has their own RSL, a physical test against their limits on a pre-production sample is the only thing that proves conformance.
How long does a test report stay valid?
Most retailers treat apparel chemical test reports as current for around 12 months. The practical trigger for re-testing is not the calendar but change: a new dye lot, a new yarn spinner, a new trim supplier or a change of finishing recipe all invalidate the previous report regardless of its age.
What happens if the bulk fails after production?
It depends entirely on what failed. pH drift or formaldehyde from a finishing bath can sometimes be corrected by re-washing and re-testing, which costs time but not the order. Nickel release from a metal trim or a banned amine in the dyestuff cannot be fixed on finished garments — those trims or that yarn has to be replaced, which is effectively a re-make.