In 2013, three researchers at Oregon State University did something most home cooks never would: they simmered tomato sauce in a stainless steel pan for six hours, then sent the sauce to a mass spectrometer to find out what the pan had given up.1
Kristin Kamerud, Kevin Hobbie and Kim Anderson ran four commercial sauces through the same procedure, cooking cycle after cycle in new stainless. After six hours, the nickel in the sauce had risen as much as twenty-six-fold and the chromium as much as seven-fold against where they started. By the tenth consecutive batch — a pan well past its first use — each 126-gram serving still carried, on average, 88 micrograms of nickel and 86 micrograms of chromium that had not been there before it met the steel.1 The new pan leached the most. The numbers fell as the surface aged, but they never reached zero.
This is worth sitting with, because of which pan it is. It is not the non-stick pan with the coating everyone has learned to distrust. It is the bare stainless one — the honest, plain-metal pan a careful shopper retreats to precisely because it feels like the surface with nothing to hide. One of the clearest measured migration records in the whole category belongs to the safe choice.
That inversion is the centre of this report, but it is not the whole of it. The deeper problem is that a buyer standing in a shop has no way to find any of this out from the thing they are about to buy. The word on the box — titanium, ceramic, 5-ply, surgical, non-toxic — is doing the informing, and the word names the wrong layer. It names a metal, a construction, a grade, or a coating's hardness. It never names the one variable that decides what ends up in your food: what actually migrates off the surface that touches it.
A pan is a stack, and the word sits on the wrong layer
Start with what a pan physically is. Almost any modern pan is not one material but several, bonded together. There is a body — usually aluminium, for how fast it spreads heat. On a coated pan there is a food-contact surface laid over that body. On a clad stainless pan there are alternating layers of steel and aluminium pressed into a wall. The food touches exactly one of these. The marketing word, more often than not, describes one of the others.
Take titanium. The word can sit on a pan in two quite different ways, and neither way tells the buyer what comes off the surface. Our Place sells a "Titanium Always Pan Pro" whose product page markets a coating-free titanium interior — what it calls "NoCo" technology, an interior it describes as developed entirely without coatings — and lists PFAS and PTFE separately among the things the pan is made without.2 Read carefully, the word titanium there is naming the interior material itself, and a good part of the page's work is telling the buyer what the pan is without. Even so, an honest coating-free titanium interior is still a surface that food sits against for hours of acidic simmering, and the page does not tell the buyer what is known about what migrates from it.
That is one use of the word. In some "titanium" non-stick, the word does something else entirely: "titanium" refers to titanium compounds added to a fluoropolymer coating to make it harder and more scratch-resistant — a reinforcement worked into the surface, not a solid-titanium cooking layer. That titanium reinforcement can be built into a PTFE coating or into other coating systems; either way the titanium is a property of the coating, not the surface your eggs meet, and certainly not a titanium pan. The two cases sit at opposite ends — a genuine coating-free interior versus a fluoropolymer reinforcement — and yet they share the same blind spot. Whether titanium names what the interior is or what the coating is reinforced with, the word answers "what is it" or "what is it not". It never answers "what comes off the surface into the food."
Take 5-ply. Viking sells a stainless line under the collection name "5-Ply Professional".3 "Five-ply" is an honest description — it names five bonded layers in the pan's wall, a construction that spreads heat evenly and resists warping. But here is the thing a buyer is not told: the food in a five-ply pan touches one layer, the inner stainless face, and what migrates depends on that surface — its steel grade, the acidity of the food, and the time spent against it — not on how many plies are stacked behind it to spread the heat. Ply count describes the wall's construction. It says nothing about what moves out of the surface the food actually meets. The word is accurate and beside the point.
Surgical and T-304 work the same way. "Surgical grade" has no defined meaning for cookware — there is no cookware-specific standard behind it; the formal "surgical" stainless standards that exist govern surgical instruments, not frying pans — and "T-304" names a steel grade: roughly 18 per cent chromium and 10 per cent nickel, the austenitic stainless that gets called 18/10. That grade is real, and for a bare stainless pan it genuinely is the metal the food meets. But 304 stainless is the exact composition Kamerud's team watched leach nickel and chromium into the sauce.1 The grade word tells you what the surface is made of. It does not tell you that the surface gives some of itself up, and it does not put a number on how much.
None of these words is a lie. That is the uncomfortable part. 5-ply honestly names construction; T-304 honestly names a grade; titanium honestly names either an interior or a reinforcement. The problem is narrower and harder to shake: no word on any of these boxes — accurate or not — tells the buyer what migrates from the surface their food actually touches. The migration variable rides on no word and no label, for any surface in the category.
What the coating actually is, underneath the word
The coated side of the category has its own version of the same gap, and it starts with how many layers there are.
A non-stick surface that the box calls a single "coating" is, in a reinforced system, not one thing but a small stack of engineered layers cured onto the metal. An industry datasheet for one such reinforced system describes it as "a unique blend of PTFE, PFA, and FEP combined with ceramic fillers" and notes that "for additional durability, ceramic fillers have been added to the primer and midcoat layers."4 Unpacked, that is three fluoropolymers and a mineral filler. They are built up in sequence — a primer, a pigmented mid-coat, and a top-coat — with only the top one ever touching food. This is how reinforced non-stick is built as a category. It is documented in coating-house datasheets and in granted patents for fluoropolymer non-stick systems.4,5 What it is not, anywhere, is printed on the box. The buyer sees "coating", or sees "titanium", and the stack underneath stays a category fact rather than a label fact.
The fear attached to that stack is PFAS — the per- and polyfluoroalkyl substances that include the PTFE in the coating. And here the honest finding cuts against the intuition. The clearest documented acute hazard of an intact non-stick coating is overheating, not a slow leach into your dinner. A medical reference on polymer fume fever notes that PTFE releases fumes from about 260°C, but that the symptoms in people generally do not begin until the coating is heated to around 350°C.6 The coating itself begins to break down in earnest — pyrolysis — at about 400°C.6 Those are temperatures a pan reaches when it is left empty on a high flame, dry, with nothing in it to cap the surface temperature. Put food or water in the pan and the surface stays far below that line. The real, documented non-stick hazard is the empty pan forgotten on a lit ring, not the coating quietly shedding into a simmer.
That should be reassuring, and partly it is. But it comes with a caveat that is itself the finding: how much an intact coating sheds into ordinary food over its service life is thinly measured in the public literature. Some food-simulant and cookware PFAS-migration studies do exist; what is missing is a standardised assay using real acidic foods across a coating's ordinary service life, measured the way Kamerud measured the steel. The metals dataset is large and precise. The intact-coating-migration dataset is thin. That asymmetry is real, and it matters — not because it proves the coating is leaching, and not because the limited data proves it is inert, but because it means the two surfaces a buyer is choosing between have been measured to wildly different depths. You can put a number on what the steel gives up. You cannot, yet, put the same kind of number on what the intact coating gives up under real cooking. The box tells you neither.
Why "how much" is the wrong question for one surface and the right one for the other
There is a second axis here, and it is the one that decides what actually matters — and no label expresses it either.
On 17 September 2020, the European Food Safety Authority published an opinion that lowered the tolerable intake for PFAS dramatically. It set a group tolerable weekly intake of 4.4 nanograms per kilogram of body weight — with effects on the immune system judged the most critical concern.7 That limit covers four substances added together: PFOA, PFNA, PFHxS and PFOS. Nanograms. Billionths of a gram. The figure is that small for a specific reason: these substances persist. The body does not readily clear them, so they accumulate over years, and a tolerable intake has to be set against a lifetime of build-up rather than a single meal.
Set that against the metals. The same agency gives dietary nickel a tolerable daily intake of 13 micrograms per kilogram of body weight, derived through benchmark-dose modelling from reproductive and developmental effects seen in animal studies — not from the skin reaction most people associate with nickel.8 That skin reaction is a separate concern. It is systemic contact dermatitis, a flare that acute nickel exposure can trigger in people already sensitised to the metal. It is real, but it is not the basis of the 13-microgram figure. The thing that matters for nickel is how the body handles it: nickel is poorly absorbed and most of what arrives is excreted, so for nickel the relevant question really is how much arrived. Aluminium behaves differently. It carries a tolerable weekly intake of 1 milligram per kilogram of body weight.9 Micrograms and milligrams — thousands to millions of times larger than the PFAS figure. Aluminium is poorly absorbed too, but the fraction that does get absorbed can be retained and build up over time. That retention is partly why its limit is written weekly rather than daily. It is not as pure an in-and-out metal as nickel, though it is still far less persistent than PFAS.
It is tempting to read those two numbers side by side — nanograms for the coating chemistry, milligrams for the metal — and conclude that non-stick is thousands of times worse. That conclusion is not in the evidence, and it is worth saying so plainly. The figures are not a severity ranking. They are anchored to entirely different endpoints — immune effects for PFAS, reproductive and developmental effects for nickel — derived by different methods, and the PFAS limit is a total-diet figure dominated by water and food generally, not by cookware coatings, whose contribution is the very thing that is sparsely measured. The small number reflects persistence and a whole-diet basis, not a thousand-fold-greater danger from a frying pan.
What the contrast actually shows is that the two surfaces fail the buyer on two different questions. For the steel, the meaningful question is how much migrates, because the body clears most of it. For the coating chemistry, the meaningful question is does it stay, because the danger is accumulation, not dose-per-meal. A buyer instinctively asks "how much leaches?" of everything — and that is the right question for the metal pan and the wrong one for the persistent chemistry. The label answers neither.
This investigation continues below.
Want the next one in your inbox?
That persistence question is exactly why PFAS sit under one of the broadest chemical restrictions Europe has proposed. On 7 February 2023, the Netherlands, together with Denmark, Germany, Norway and Sweden, published a REACH proposal covering thousands of PFAS substances across many uses.10 That proposal is a class-wide measure spanning far more than cookware — it is not a verdict on any one pan — but it is there because these substances last, which is the same reason their tolerable intake is written in nanograms.
The law asks the maker to tell the seller, not the seller to tell you
So why can a buyer not simply find the contact layer printed somewhere? Because the disclosure duty, where it exists, points the wrong way for the shopper.
A buyer in Great Britain buys their pans under GB law. After the Brexit transition ended on 31 December 2020, the European food-contact rules were retained as assimilated law in GB. Two of them carry the framework. Regulation (EC) 1935/2004 sets the general safety and traceability rules for anything that touches food.11 Regulation (EU) 10/2011 governs plastics and fixes the overall migration limit.12 In England, both of those rules are enforced through one domestic instrument: the Materials and Articles in Contact with Food (England) Regulations 2012.13 The Food Standards Agency is the central competent authority that administers and oversees the regime. The day-to-day enforcement is carried out by local authorities — trading standards and environmental health officers. The FSA's own guidance is clear that regulated food-contact materials "need to be authorised before they can be used and placed on the market in Great Britain" and must comply with assimilated 1935/2004.14
All of that is a genuine safety gate, and it sits upstream of the shelf — but it does not work the way "authorised before it reaches a shop" might sound. Specific regulated food-contact substances and categories require authorisation, and every food-contact material must comply with the general safety framework; what stands behind a pan is not an individual pre-clearance of that pan but compliance with those rules, evidenced through supply-chain paperwork. The central document is the Declaration of Compliance, which the maker passes to the seller. It is never a layer list printed for the buyer. There is no consumer-facing requirement, in GB, for a cookware box to name the primer, the mid-coat, the top-coat, or the steel grade the food actually contacts. A UK buyer has no legal right to a printed layer list, and that absence is not an oversight to be indignant about — it is how the system is built. The compliance evidence runs business-to-business; the buyer is simply not its audience.
That is the vacuum the marketing word fills. The law declines to require the box to say which layer meets the food and what comes off it, and into that silence steps titanium, ceramic, surgical, 5-ply — words that sound like answers to a question they do not address. The word does the informing the law does not require, and it informs about the wrong layer.
It is worth being fair to the strongest version of the opposing case, because it is right about a great deal. No manufactured product lists its full bill of materials: a car does not name its brake-pad binder, a kettle does not declare its element alloy. The migration limits are real and enforceable, and for a typical adult both surfaces, used normally, sit within the relevant intakes. But the two cannot be expressed the same way. Kamerud's tenth-cycle 88 micrograms of nickel per serving sits below EFSA's daily nickel tolerable intake once an adult's body weight is reckoned in.1,8 The PFAS picture cannot be put in the same sentence. EFSA's PFAS limit is a total-diet figure, and the cookware coating's contribution to it is precisely the thing that has not been well quantified. An intact coating with food in the pan never approaches its fume threshold,6 but that is a heat fact, not a migration figure. The buyer is genuinely protected. Every one of those points is true. And not one of them lets the buyer answer the question they are actually asking when they pick the pan up: which layer meets my food, and what is known about what comes off it. The protection works without informing. The complaint here is not that the pan is unsafe. It is that the information needed to choose between two surfaces lives everywhere — in the datasheet, in the Declaration of Compliance, in the EFSA opinion — except at the point of choice.
What would actually settle it
There is one dataset that would convert this whole picture from an asymmetry into a clean comparison, and it is the one that is missing. Limited food-simulant and cookware PFAS-migration studies exist, but there is no published, standardised assay quantifying how much PFAS migrates from an intact non-stick coating into real acidic food across a coating's service life, under ordinary home cooking — nothing comparable, on the coating side, to Kamerud's stainless protocol. Produce those numbers and set them beside Kamerud's stainless figures, measured the same way, and a buyer could finally rank the two surfaces on a single axis. Right now the steel is measured to the microgram and the intact coating is not measured on those terms, which is precisely why the honest finding is an asymmetry of knowledge rather than a verdict on which pan is worse. That missing assay is the thing that would update this assessment — and its absence is the cleanest evidence that the buyer is being asked to choose between two surfaces that have not been measured on the same terms.
What this changes for Saturday
The practical upshot is calmer than the chemistry might suggest, because the variables that drive exposure are conditions, not hidden poisons — and conditions are something a cook controls.
For stainless, the thing that actually moves nickel and chromium into food is the steel grade and the combination of acid plus time, not the ply count or the word "surgical". So the useful question to ask a brand is the steel grade — 18/10 versus 18/0, which is the difference between a nickel-bearing surface and a nickel-free one, and nickel is the migration variable that matters most, especially if anyone in the household has a nickel sensitivity. And the useful habit is to keep long, acidic simmers — a six-hour tomato ragù — off a brand-new stainless pan, since new steel leaches the most and the rate falls with age.1
For non-stick, the lever is heat, and the documented hazard is the empty pan overheating. So the rule is simple: do not preheat a coated pan empty on a high flame. Food and water cap the surface temperature near boiling, but oil alone — or a high-heat sear — can still climb toward the fume range, so cook on low-to-medium heat and don't let the oil smoke.6 And when a box says "titanium" or "ceramic", read it as a claim about hardness, or about what the surface is not (PFAS-free, where that is stated), not as a description of what comes off the surface your food meets — because the box does not describe that, and now you know why.
If you want to step off the coating question entirely, the genuinely long-lived options are the old ones: cast iron and carbon steel, which build their own slick surface with use and last for decades, and enamelled cast iron — a Lodge skillet, a Le Creuset casserole — where the food-contact layer is a fired glass enamel rather than a fluoropolymer or a bare reactive metal. None of these is sold by us, and none is a perfect answer: cast iron is heavy and needs seasoning, enamel can chip at high heat, and all of them carry their own trade-offs. But each one names what your food touches, which is more than the reassuring word on the premium non-stick box manages to do.
The deeper fix is not a better word. It is a box that names the contact layer — the steel grade, or the actual top-coat polymer — and a migration number measured the way Kamerud measured the metals. Until a label does that, the most honest thing on the shelf is the cast-iron skillet, because at least there is no word standing between you and the surface.