play · health · circular-economy
hand-me-downs from before the ban
Chemical limits for toys apply when a toy is first sold, not when it is passed on: surveys in Japan and the UK found second-hand plastic toys made before the limits still holding restricted chemicals, some above today's limits, and toy libraries and resale shops have no tested low-cost way to tell which ones
Source: Bekki et al., “Comprehensive survey on the use of plastic additives in toy…” Open ↗source: primaryLast updated: 2026-11-17
Problem statement
Babies and toddlers put toys in their mouths, so rich countries limit what toy plastics may contain. Japan has limited three phthalate plasticisers to 0.1% in toys for children under 6 ("designated toys") since 2010, and restricts others in toys meant for the mouth. The EU's 2009 Toy Safety Directive sets limits on how much cadmium, lead and other elements may leach out. Those limits apply when a toy is made and sold. Many toys, though, are handed down, swapped, lent by toy libraries or resold. A 2024 survey led by researchers at Japan's National Institute of Public Health analysed 84 plastic toys bought online, in toy shops, in supermarkets and from second-hand dealers in Saitama. It found regulated phthalates (DIDP, DNOP, DINP) above the 0.1% limit "in used products manufactured before regulations were passed in Japan". The paper adds that the toys holding them were outside the rule: the one with the most DIDP (up to 62%), a 1990s finger puppet bought from a second-hand dealer, "was not initially intended to come in contact with the mouths of infants", and the toys over 0.1% for DNOP and DINP "were not targeted by the regulations". The authors conclude that "the safety management of used products is crucial". A 2018 UK study of 197 second-hand plastic toys from two nurseries, a primary school, charity shops and five family homes in Plymouth tested 34 components of 26 of them in simulated stomach acid. Cadmium or lead leached above the current EU limits in eight cases: components of four toys (red and yellow bricks from two Lego sets, a PVC model dinosaur and a PVC tape measure), with two more toys potentially non-compliant on chromium. The paper notes that "there is no retroactive regulation on second-hand toys". In the UK, second-hand toys are covered by the general product safety rules rather than the toy regulations. They "must still be safe", but the trade guidance's practical check is a CE or UKCA mark, which shows whether the toy was likely legal when new, not whether it meets today's limits (Business Companion). The unsolved problem sits at the point of reuse. A parent, a toy-library volunteer or a charity-shop sorter has no tested, affordable way to tell which old plastic toys are over today's limits. Colour-change home test kits failed a US regulator's evaluation, and the methods that work (mass spectrometry for phthalates; X-ray fluorescence, then acid migration tests, for metals) are laboratory or professional tools.
Why this matters
The children most exposed are the youngest, who mouth toys most. The Japanese authors stress this "mouthing" route and note that plasticisers also move into indoor dust. Reuse is where old toys concentrate. In the Japanese survey, products made in China before 2010, and Chinese-made products with no date, clustered together in the chemical analysis, and the authors suggest the undated ones "might have been manufactured before 2010". DEHP, banned in designated toys since 2002, was much lower in products made after 2010. So the rules worked for new toys, and the over-limit phthalates the survey found were in the stock that predates them. The UK study found bromine (a marker for brominated flame retardants), cadmium and lead often in second-hand toys, at maximum concentrations of about 16,000, 20,000 and 5,000 μg/g. Its author notes that old plastic toys "often appear to be in good condition" and so tend to be reused more than other old toys, that they remain in nurseries, hospitals and schools, and that the UK actively encourages toy reuse. Lead from old electronic plastics and pigments (not from PVC) is also carried forward into newer products through recycling (Turner and Filella 2021).
What’s been tried and why it hasn’t worked
Limits at the point of sale (successful for new toys). Japan's designated-toy limits reduced regulated substances in toys made after they came in: the Japanese survey shows DEHP much lower in products made after 2010. (No before-and-after data for the EU directive was read.) But by design they do not reach toys already in homes, and the UK study states there is no retroactive rule for second-hand toys. The Japanese survey also found substitutes: ATBC in 52% of products, up to 48% by weight of the chemicals in PVC balls. The authors note that its toxicity "might be unknown", so passing a phthalate limit is not the end of the question.
Laboratory screening (works, but not where toys change hands). Both studies needed laboratory methods: liquid chromatography–mass spectrometry and GC-MS in Japan; X-ray fluorescence plus simulated-stomach-acid migration and plasma spectrometry in the UK. Portable spectrometers sold to testing labs and inspectors exist, but none read here is aimed at, or priced for, the places where toys are passed on.
Handheld X-ray fluorescence screening (works for metals, but is a professional tool). Handheld XRF has been used on toys already in shared use: a US study tested 535 toys from 10 licensed day-care centres in Las Vegas and found 29 over 600 ppm lead, of which 20 were PVC and 17 yellow (Greenway and Gerstenberger 2010). The UK study used XRF to pick which toys to send for migration testing, and treated a total concentration above a migration limit as "a trigger for further investigation". Two limits remain. First, total content is "not, necessarily, a good proxy for exposure", since total and migratable amounts correlated only for cadmium and lead. Second, the US Consumer Product Safety Commission found XRF correctly identified 12 of 13 lead-containing samples, but noted that "XRF detectors are generally not available for consumer use" and need "knowledge, training and consideration of its limitations".
Home test kits (tested, and failed). In 2007 the CPSC evaluated common colour-change (rhodizonate and sulfide) lead test kits on paints and other products: "Of 104 total test results, more than half (56) were false negatives, and two were false positives." The kits were designed for the high lead levels of household paint and "may not be useful ... for detecting lead in other materials, such as metal jewelry or vinyl products". The agency advised that "consumers should not use lead test kits to evaluate consumer products for potential lead hazards".
Labelling and advice to consumers. The UK study considers a "right to know" labelling approach and judges it "difficult to implement" for old toys, "because of the plethora and variety of old toys on sale or passed down whose precise ages and origins are unknown". Its own recommendation is awareness of "small, mouthable, and brightly colored (and in particular red and yellow) old plastic toys". That colour cue is useful but imperfect in the paper's own data: 1990s Lego bricks "visually indistinguishable" from cadmium-bearing 1970s bricks contained no detectable cadmium, and nearly 40 non-green, mostly neutral-coloured items carried bromine, a flame-retardant marker.
What would unlock progress
A triage step cheap enough for a toy library, a nursery or a resale shop. It should flag the toys most likely to be over the limits for retirement or for lab testing, and pass the rest. It has to cover two different hazards. One is soft PVC (phthalate plasticisers, and lead or cadmium stabilisers in older items). The other is brightly coloured hard plastic from before the limits (cadmium and lead pigments, as in the old Lego bricks). Cues exist separately. Manufacture date separates high and low phthalate content in the Japanese data. PVC and yellow colour predicted high lead in the Las Vegas day-care toys, and red and yellow colour flagged cadmium in the UK toys. XRF can tell PVC from other plastics by its chlorine (both the Las Vegas and UK studies did this), but the Las Vegas authors saw no visual or tactile difference between PVC and non-PVC toys, and no visible difference between yellow toys with high lead and yellow toys with no detectable lead. But each cue misses cases. Date did not separate lead in consumer goods in Turner and Filella's data: lead was detected in 11.9% of items made both before and after the 2006 electronics rule, which the authors attribute in part to recycled plastic carrying it into newer goods. Colour missed the bromine in neutral items and could not tell old Lego from newer Lego. Nobody read here has combined the cues into a tested rule or tool, or measured how well such a rule sorts toys against lab results.
How we’d know it’s being solved
- Sensitivity and specificity of a triage rule or device against laboratory results on the same toys: the share of over-limit toys it flags (missing few matters most) and the share of compliant toys it passes.
- Cost per toy screened and time per toy, against what a volunteer-run toy library can bear.
- Use: whether intake volunteers apply it consistently. This can be measured as agreement between two volunteers on the same batch.
- A proxy to avoid: "toy looks old" or "has no CE mark" on its own. Neither is a measurement of content. The trade guidance treats a missing mark as a sign that "it is likely that they were not legal when they were new". A mark that is present does not show the toy meets today's limits: the EU cut its cadmium and lead migration limits after the 2009 directive (UK study, Table 1; the lead limit for scraped-off material to 23 mg/kg in 2017, per Turner and Filella).
Where to start
Facility-free door (published data and public rules only). Write a one-page intake rule for second-hand toys for under-3s from cues that a volunteer can see: date of manufacture where it can be found, type of toy (the UK study found most of its PVC among figures, such as animals, dinosaurs and dolls, and among bath toys; the rule must not ask volunteers to judge the plastic itself, because the Las Vegas study found PVC toys looked and felt no different from the others), colour (bright red, yellow or orange), and whether the toy is small or made to be mouthed. Then score it on paper against the item-level data that is already published. The UK study's Table 3 lists 30 toy components with their colour, PVC or not, and migration results, with non-compliant ones marked. The table holds only components in which at least one element was detected in the extract, so it over-represents problem toys; report the scores with that caveat. Score how many of the non-compliant components the rule would have flagged and how many compliant ones it would have retired. Then check whether the Japanese survey's supplementary Table S1 and the UK study's Table S1 give item-level dates, resins and results, and extend the test if they do. Output: the rule, its scored test, its known misses (neutral-coloured items carrying bromine; new items with recycled lead), and a list of what data would be needed to test it properly.
Make door (teaching engineering lab): a low-cost "soft PVC, send for testing" screen. Start from what exists. Colour-change lead swabs are not a starting point: the CPSC found more than half their results were false negatives on paints and other products. Handheld XRF works for metals and reads chlorine as a PVC marker, but it is a trained professional's tool. Portable spectrometers are reported to be marketed to testing labs as phthalate prescreens (known here only from a search summary of vendor material). The team should check and price these before claiming a cost gap. Build a handheld triage aid that tests whether low-cost near-infrared or similar optical sensors can sort toy plastics into "likely soft PVC" and "other", paired with the colour, date and mouthing intake rule. It is a plastic-type sorter, not a chemical detector: it cannot say whether a toy holds lead, cadmium, bromine or phthalates, and some over-limit toys are not PVC at all (the cadmium-bearing Lego bricks in the UK study's Table 3). It triages and does not certify: anything it flags goes to retirement or to a lab. It must be non-destructive, read only from the toy's surface. No scraping, sanding, grating, cutting, acid extraction or heating of toys outside a supervised laboratory. Old toys may contain lead, cadmium or brominated flame retardants, so anyone handling them, in bench tests or user sessions, wears disposable gloves, does not eat or drink during the session, keeps toys bagged between tests, and washes hands afterwards. Bench-test it on donated toys and on reference plastics of known type. Test users: adults who sort or pass on second-hand toys, such as volunteers at a campus childcare centre's toy exchange or a community toy library, charity-shop sorters, and parents, in a supervised session with toys only. No children take part, and no prototype leaves the session. Access line: checking the device against true content needs a partner lab (mass spectrometry for phthalates; XRF and EN 71-3 migration testing for metals, run by the lab). If a university department holds a handheld XRF, its trained operator, not the students, would run it. Without a partner, the team can test only plastic-type identification against reference samples.
Design-the-study door. Design a sampling study a public-health lab or consumer body could run: collect a stratified sample of second-hand mouthable toys from toy libraries and resale outlets, run the triage rule blind, and lab-test all items. Include a pre-registered estimate of sensitivity and specificity. The team designs it and does not run it.
Genes
Related problems
Stuck the same way, in other fields.
field: water & energygreen by paperwork, not by physicsGreen hydrogen certification temporal matching gapthe rules got in the waysource: primaryfield: the built worldsame slab, four rulebooksA reclaimed hollow-core slab is waste in Finland, a product in the Netherlands and ambiguous in Sweden — under the same EU directive — and no standard says how many slabs to test before re-using a building's worththe rules got in the waysource: primaryfield: health & medicinethe app was prescribed, the outcome unknownReal-world outcome measurement for prescription digital therapeuticsthe rules got in the waysource: secondaryfield: digital & spacecheap rockets, crowded launchpadsLaunch site throughput and range operations bottleneckthe rules got in the waysource: secondarySources
- Bekki K, Eguchi A, Takaguchi K, Inaba Y, Yukawa K, Yoshida S, Azuma K. "Comprehensive survey on the use of plastic additives in toy products used in Japan." Environmental Health and Preventive Medicine 29:43, 2024, doi:10.1265/ehpm.24-00054, accessed 2026-11-17. Link ↗
- Turner A. "Concentrations and Migratabilities of Hazardous Elements in Second-Hand Children's Plastic toys." Environmental Science & Technology 52(5):3110–3116, 2018, doi:10.1021/acs.est.7b04685 (author's accepted manuscript in the University of Plymouth repository PEARL, handle 10026.1/11188, read in full via the Internet Archive copy), accessed 2026-11-17. Link ↗
- Turner A, Filella M. "Lead in plastics – Recycling of legacy material and appropriateness of current regulations." Journal of Hazardous Materials 404:124131, 2021, doi:10.1016/j.jhazmat.2020.124131 (published version, open access, read in full from Archive ouverte UNIGE), accessed 2026-11-17. Link ↗
- U.S. Consumer Product Safety Commission. "CPSC Staff Study: Home Lead Test Kits Unreliable." News release, October 22, 2007 (read via the Internet Archive copy of 2026-08-09), accessed 2026-11-17. Link ↗
- Greenway JA, Gerstenberger S. "An Evaluation of Lead Contamination in Plastic Toys Collected from Day Care Centers in the Las Vegas Valley, Nevada, USA." Bulletin of Environmental Contamination and Toxicology 85(4):363–366, 2010, doi:10.1007/s00128-010-0100-3 (journal article read in full through library access), accessed 2026-11-17. Link ↗
- Chartered Trading Standards Institute, Business Companion. "Toys" (last updated June 2026), accessed 2026-11-17. Link ↗
Revision notes
This brief gets checked against its sources and contains the following revisions. Future revisions get added here.
Worth knowing
- Only the UK study's cases exceed limits that apply to those toys; the Japanese toys above the phthalate limit were outside the rule's scope.
- No figure was found for how many toys are passed on second-hand, so the page makes no claim about the size of that trade.
Full check
Sources read
- Bekki et al., plastic additives in toys used in Japan (Environmental Health and Preventive Medicine, 2024): full text
- Turner, hazardous elements in second-hand plastic toys (Environmental Science & Technology, 2018): full text (the author's accepted manuscript)
- Turner and Filella, lead in recycled plastics (Journal of Hazardous Materials, 2021): full text
- US Consumer Product Safety Commission, home lead test kits release (2007): full text
- Business Companion (Chartered Trading Standards Institute), toys guide: full text
- Greenway and Gerstenberger, lead in day-care toys in Las Vegas (Bulletin of Environmental Contamination and Toxicology, 2010): full text
What changed
- Stated more strongly than the source: the Japanese toys over the phthalate limit: the page now adds the paper's point that these toys were outside the rule's scope (one was a 1990s finger puppet not made for the mouth); the home lead kits' failures were said to be on consumer products; the regulator tested them on paints and other products; colour was said to be unable to tell old Lego from new; the study compared 1970s bricks with 1990s ones, so the page now says newer Lego.
- Contradicted by the source, corrected: in the build starting point's access line, the Japanese survey's method for phthalates: it used mainly liquid chromatography with mass spectrometry, so the page now says mass spectrometry.
- Title: the long title no longer says the old toys were still over the limits; it now says they still hold restricted chemicals, some above today's limits (display title unchanged).
Left open
- that portable spectrometers are marketed to labs as phthalate prescreens [known only from a search summary of vendor material; the page says so]
Separate check of specific claims: Two related pages were named; no fact on the page changed.
Earlier full check: Eleven details were corrected, including a citation, who led the Japanese survey, what a CE or UKCA mark shows, and the free starting point, which had asked volunteers to tell soft PVC by sight although the source says it looks and feels no different, and now uses the type of toy instead.
Separate check of specific claims: The build starting point now says plainly that it sorts plastic type and does not detect chemicals, and adds handling rules (gloves, no sanding, hand-washing); the free starting point notes that the published table it uses over-represents problem toys.
Earlier full check: Both UK papers were read in full, so the UK results are now given as parts of four toys; the page adds that home colour-change kits failed a regulator's test, that handheld X-ray screening is a professional tool and that UK guidance says second-hand toys must still be safe; the long title now says no tested low-cost check exists (display title unchanged).