Toxic Chemicals in Mattresses: 21 Found in 16 Children’s Beds Tested

Almost everything written about chemicals in mattresses has one weakness in common: it tests the mattress sitting still, at room temperature, with nothing on it. That is a reasonable way to check a product against a regulation and a poor way to describe what a child is exposed to. A sleeping four-year-old is a warm, heavy object lying on a compressible foam slab for half of every day, and both the warmth and the weight change the chemistry of what comes off the surface.

Two papers published in 2025 by a group at the University of Toronto, led by Samira Vaezafshar with Miriam Diamond’s laboratory, closed that gap from both ends. One bought new mattresses and measured them under realistic sleeping conditions. The other went into twenty-five real family bedrooms and sampled the air both around the room and in the narrow layer of air immediately surrounding a sleeping child. Read together, they give the clearest picture yet of what a bed actually emits, and they are specific enough to act on.

What chemicals are in mattresses?

Four families, added for four different jobs. The study is “Are Sleeping Children Exposed to Plasticizers, Flame Retardants, and UV-Filters from Their Mattresses?” by Vaezafshar, Wolk, Simpson, Akhbarizadeh, Blum, Jantunen and Diamond, in Environmental Science & Technology, volume 59, issue 16, pages 7909–7918, 2025. Sixteen new children’s mattresses were bought at ordinary retail, their covers and foam cores analyzed separately, and the cover polymers identified by infrared spectroscopy. Twenty-one compounds were detected across the set.

Chemical family	Why it is in a mattress	Examples detected	Where it sat
Ortho-phthalates	Plasticizers — they keep a waterproof cover soft and flexible instead of stiff	DnBP, DiBP, BzBP, DEHP, DnOP, DiNP	Mostly the cover
Organophosphate esters	Flame retardants, added to foam to meet flammability rules	TCEP, TCPP, TDCiPP, TEHP, TBOEP, TPhP	Foam core and cover
Benzophenones	UV filters that stop dyes and polymers fading or degrading in light	BP, BP-3, BP-8	Cover
Salicylates	Also UV absorbers, used alongside or instead of benzophenones	2-EH-salicylate, phenyl salicylate, methyl salicylate	Cover

One result in that table deserves emphasis, because it disposes of the most common shopping heuristic. The researchers found no consistent relationship between the cover’s polymer type and what was in it — rigid PVC covers were not reliably worse than flexible polypropylene-polyethylene ones. You cannot read the risk off the material name on the tag.

How much was actually in them?

Enough, in several cases, to cross a legal line. Canada caps di-n-butyl phthalate in children’s mattresses at 0.1% by weight. These are the measurements that stand out from the sixteen:

Chemical	Highest measured	Regulatory position
DnBP (phthalate)	0.22% — 2,200 mg/kg in one cover	Over the 0.1% Canadian limit for children’s mattresses
DiNP (phthalate)	0.27% — 2,700 mg/kg in one cover	Capped in children’s toys; not capped in mattresses
DiBP (phthalate)	0.24% — 2,400 mg/kg in one cover	Capped in children’s toys; not capped in mattresses
DnOP (phthalate)	1,700 mg/kg in one cover	Capped in children’s toys; not capped in mattresses
TCEP (flame retardant)	9,000 mg/kg in one foam core	Prohibited in Canada since 2014
TEHP (flame retardant)	31,000 mg/kg — about 3% by weight of one foam	The highest single concentration in the study
TBOEP (flame retardant)	22,000 mg/kg in one cover	Detected in 88% of the mattresses tested

Five of the sixteen mattresses carried organophosphate flame retardants at between 1% and 3% of their weight. The authors made a pointed observation about that: Canada’s flammability standard for these products can be met without adding them at all, so in those five cases the chemistry was not doing a job the law required. And the TCEP result is the one that says most about oversight, since the compound has been off-limits in Canada for a decade and turned up at a level the paper described as higher than all previously published data.

There is a second, quieter finding worth knowing before you shop. The team bought two mattresses of the same brand and model roughly two years apart. The 2021 sample had a PEVA cover and the highest DnBP reading in the study; the 2024 sample had a polypropylene cover, no detectable DnBP — and more DiNP. The market moved off one phthalate and onto another. A certification checked once is not a guarantee about the box that arrives.

Why does a sleeping child get a bigger dose?

Because heat and pressure both drive emission, and a child supplies both for ten hours at a stretch. Eight of the mattresses were put through a three-stage emission test using passive samplers on the surface: room temperature; a child’s body temperature of 37.5 °C; and body temperature with 7.5 kg of weight applied to compress the foam the way a sleeping body does.

Test condition	Compounds emitted	Median TCPP emission
Room temperature, no weight	12 of 45 measured	74 pg/cm²/h
Body temperature (37.5 °C)	20	380 pg/cm²/h
Body temperature plus 7.5 kg weight	21	—

The individual jumps are larger than the headline. TCPP emissions rose roughly five-fold when the surface was simply warmed. TBOEP went from below the detection limit with no weight on the mattress to a median of 460 pg/cm²/h once weight was applied, and its detection frequency went from three of the eight mattresses to all eight. DiNP was detectable coming off one mattress in eight at room temperature and four in eight under heat and weight, with the median rising from below detection to 360 pg/cm²/h.

That is the practical lesson of the whole paper. A mattress that emits almost nothing in a warehouse can emit all twenty-one compounds under a warm, sleeping child. Any test or claim that does not say what temperature and load it used is describing a different object.

Is the air over the bed worse than the rest of the room?

It measured worse, in real homes. The companion paper — “Young Children’s Exposure to Chemicals of Concern in Their Sleeping Environment: An In-Home Study,” Vaezafshar, Wolk, Arrandale, Sühring, Phipps, Jantunen and Diamond, Environmental Science & Technology Letters, volume 12, issue 5, pages 468–475, 2025 — put passive samplers into 25 bedrooms of children aged 6 months to 4 years in Toronto and Ottawa, and sampled three places at once: the general bedroom air, the thin layer of air immediately around the sleeping child, and the mattress surface.

Sampling location	Compounds detected (of 51 targeted)
General bedroom air	28
Air immediately around the sleeping child	31
Mattress surface	30

Concentrations in the sleeping layer exceeded those in the room air around it, which the authors read as elevated exposure during sleep, with mattresses identified as a source of specific phthalates and organophosphate esters. Diethyl phthalate and benzophenone were the two detected most often. Older mattresses showed higher concentrations of certain phthalates — a reminder that this is a slow release over years, not an off-gassing event that finishes in the first week.

Are these microplastics?

No, and it is worth being precise about it, because the two get run together constantly. A microplastic is a solid fragment of plastic. Everything measured in these two studies is a semivolatile chemical that was blended into a plastic rather than chemically bonded to it, which is exactly why it can leave: it was never part of the polymer chain. Phthalates and organophosphate esters migrate out of foam and film into air, dust and skin over the product’s life.

So this is a plastic-related exposure that travels a different road, and the countermeasures differ too. A HEPA filter catches particles; it does not remove a gas-phase plasticizer. Washing fabric and reducing the number of treated soft objects in the room does more here than filtration alone. Synthetic bedding and covers can also shed polyester fibers into household dust — a genuine issue we cover in microplastics in indoor air and microplastics in clothing — but fiber shedding is not what these papers measured, and we are not going to imply it was.

Do adult mattresses have the same problem?

Unknown from these studies, and we will not pretend otherwise: both papers tested children’s mattresses, and neither sampled an adult bed. What can be said honestly is that the materials are the same family — polyurethane foam cores, plasticized waterproof or quilted covers, the same flame-retardant chemistry meeting the same flammability standards — and that the two drivers the study identified, body heat and body weight, are larger in an adult, not smaller. An adult also sleeps on the same mattress for far more years than a toddler does.

That is a reasonable inference, not a measurement, and the difference matters. If someone tells you a specific number of chemicals comes off an adult mattress, ask which study measured it.

What actually reduces the exposure

The researchers’ own recommendations are unglamorous and cost almost nothing, which is usually the sign of good advice. In rough order of effect for the effort:

1. Wash bedding and sleepwear frequently. Clean fabric is a physical barrier between the child and the mattress surface, and washing removes compounds that have already accumulated in it. This was the team’s first suggestion.
2. Clear the bed. Every surplus pillow, blanket and stuffed toy is another treated textile that absorbs these compounds and re-releases them into the small volume of air the child actually breathes. Keep what is used and remove the rest.
3. Prefer undyed or neutral fabrics. UV filters such as benzophenones are added largely to protect dyes and polymers from light. Fewer dyes means less reason for them to be there.
4. Let a new mattress breathe before it goes into the room. Emission is highest when a product is new and warm. Unwrapping it in a ventilated space for a few days does not solve the long tail, but it moves the steepest part of the curve out of the bedroom.
5. When it is genuinely time to replace, buy the source out. A mattress built from materials that do not require plasticizers or added flame retardants — organic latex, organic cotton and wool over coils — removes the chemistry instead of managing it. Wool is the relevant detail: it passes flammability requirements on its own, which is how these beds avoid added retardants.

What we would not do is replace a serviceable mattress on the strength of these papers alone. The first four steps are free or nearly free and address the exposure route the study actually measured.

---

Mattresses built without the additives

These are the four picks from our existing mattress research that are relevant to this specific problem — beds whose construction avoids the plasticizer and flame-retardant chemistry the Toronto studies measured, rather than beds that simply market themselves as clean. Ratings shown are the verified buyer ratings on their listings; prices move constantly, so we do not print them.

---

The honest verdict

Sixteen mattresses is a small sample, they were bought in Canada, and neither paper measured a health outcome in a single child. Those are real limits and we will state them plainly rather than bury them. What the studies do establish is narrower and harder to argue with: these compounds are present in new children’s mattresses at percent-level concentrations, several exceed limits that already exist for other children’s products, one is outright prohibited, and emission rises substantially under exactly the conditions a child creates. The authors’ conclusion was a call for stricter regulation and better manufacturer oversight, not a call to panic, and that is the right reading.

For a household, the proportionate response is the cheap end first. Wash the bedding, clear the bed, skip the dyed novelty sheets. Replace the mattress on its own schedule, and when that day comes, spend the money on construction that never needed the additives — which is the same advice our non-toxic mattress guide reaches from the product side, and which pairs with the organic cotton sheets guide for the barrier layer and the bedroom air purifier guide for the particles filtration genuinely does catch.

---