What is vegan leather made of, and is it better?

"Vegan leather" names a category defined by an absence. It tells you nothing about what the material is.

One thing it tells you reliably: no animal died for it. What the thing is actually made of stays unsaid, and in most cases the answer is plastic. For plenty of buyers that remains the right trade, but anyone choosing it for environmental reasons rather than animal ones is owed the full picture, and the marketing rarely provides it.

What is vegan leather made of?

In the overwhelming majority of products on sale, one of two synthetics coated onto a fabric backing.

Polyurethane (PU). A polymer coating applied over a base cloth, usually polyester or sometimes cotton. Softer, more breathable and less chemically nasty to produce than the alternative. This is what most mid-market "vegan leather" is.

Polyvinyl chloride (PVC). The older option, and the worse one. Its production relies on chlorine chemistry and plasticisers, and it has a long-standing reputation as the least environmentally defensible material in the category. It still turns up in cheap goods.

If a product says "vegan leather" and nothing more specific, PU is the safe assumption.

Worth knowing before the comparison starts: many animal leathers carry a thermoplastic coating too. Pigmented and corrected-grain leathers are finished with polyurethane for the same reasons the synthetics are. The coating is a shared feature of the two materials rather than a line between them.

What about pineapple, cactus and mushroom leather?

These are real, they are genuinely innovative, and they are almost never what the marketing implies.

Piñatex. Developed by Dr Carmen Hijosa in the 1990s, made from pineapple leaf fibre, a by-product of existing fruit harvests. The fibres are mixed with polylactic acid from corn, then finished with a polyurethane resin coating for durability.

Desserto (cactus). Grown on nopal cactus in Mexico, low water demand, and a reported carbon footprint as low as 1.4 kg CO₂e per m². The company does not publicly disclose what proportion is cactus or what polymer binds it.

Mycelium (MycoWorks, Mylo). Fungal mycelium grown in trays, typically onto a cotton backing. Closest to a genuinely novel material rather than a coated one.

Apple leather. Made from pomace, the pulp left over from juicing. Commonly cited as around 50% polyurethane.

The pattern is consistent. A peer-reviewed review of leather alternatives puts it bluntly: despite the impression given in media coverage, many plant-based alternatives are neither fully bio-based, nor biodegradable, nor plastic-free. Most sit in a category that is neither recyclable nor compostable: mixtures of polyurethane and plant matter that cannot be separated back out.

The reason is physics rather than cynicism. Plant fibre alone gives way under the flexing, abrasion and moisture a shoe or a bag lives through. The coating is what makes the product last, and removing it is the unsolved problem in the field.

What about rexine, protein leather and "premium" vegan leather?

Rexine is the oldest of the three and the least mysterious. A coated cloth trademarked in 1915 by the British Leather-cloth Manufacturing Co of Hyde, near Manchester, later trading as Rexine Ltd, it was cotton or a cotton-rayon blend coated in cellulose nitrate and later PVC, then embossed with steel rollers to imitate a grain. It covered car seats, furniture and book bindings for decades, and ICI bought a substantial interest in the business in 1926. The brand outlived the company. Across India, Pakistan and parts of Britain, rexine is still the ordinary word for leatherette, which makes a rexine-versus-vegan-leather comparison a material set against itself, a century apart.

Protein leather is harder to pin down, because nobody selling it will say what it is. The clearest technical record is a US patent granted in 1980 for protein-containing artificial leather, in which the protein is collagen water-solvated from limed but untanned tannery flesh and slaughterhouse waste, then cross-linked onto a synthetic support that gains more than 30% of its weight from it. That is one method from one filing, and it settles nothing about what is inside a headphone ear cushion sold under the same name today. What it does settle is that the name carries no guarantee. Where a seller will not say what the protein came from, a vegan buyer has no grounds for assuming it came from a plant.

Premium vegan leather is the emptiest of the three. Animal leather has a formal definition (EN 15987 requires tanned animal hide) and a set of trade grades, full-grain down through top-grain and split to bonded, that carry rough meaning. Vegan leather has neither. No standard defines it and no grading system ranks it, so premium is an adjective with nothing behind it. The questions that carry information are which polymer, what share of the material is plant matter, and what the backing is.

The naming problem is being settled in law rather than in marketing. The word leather is legally protected in Belgium, France, Italy and Spain, and Portugal's Decreto-Lei n.º 3/2022, published on 4 January 2022, prohibits "vegan leather", "cactus leather" and "synthetic leather" as descriptions of non-animal materials, with fines attached. The trade bodies pushing these laws have an obvious commercial interest in them. They are also right about the part that matters here: the label describes what a material is not.

So is it better than animal leather?

For animals, unambiguously. No one was farmed or slaughtered.

The line drawn between species is hard to defend

Farming animals for their skin is already treated as indefensible in a good deal of the world, for some animals. More than twenty European countries have banned or are phasing out fur farming, starting with the United Kingdom in 2000. Chanel dropped crocodile, lizard, snake and stingray in 2018; Burberry, Mulberry, Victoria Beckham and Balenciaga followed; Selfridges stopped selling exotic skins in February 2020.

None of that reasoning has been extended to cattle, and cattle are where the volume is. Global raw cattle hide production reached 8.2 billion kg in 2022, which at an average hide weight of 27 kg is roughly 300 million animals. A mink, a crocodile and a cow are all farmed, confined and killed for a skin. Anyone who wants the ban to stop at the mink has to name the property that makes the cow different, and the usual answer is that leather is only a by-product of meat.

The carbon comparison

A 2026 meta-analysis in ACS Sustainable Chemistry & Engineering puts finished bovine leather at 187.1 kg CO₂e per kg (interquartile range 148.1 to 214.5) against 14.9 kg for vegan leather (10.1 to 19.5). Its own summary is that leather comes in more than twelve times higher.

Two things to hold on to about that number. It is one study, and the allocation choice behind it is exactly what the leather industry disputes, so read it as the strongest current case rather than a settled verdict. And its system boundary is cradle-to-gate: the authors state plainly that shipping, retail and end-of-life are excluded, and there is no use phase in the model at all.

How much of the cow does the hide pay for?

That allocation choice is the argument. Most industry emissions figures rest on the European Commission's Product Environmental Footprint Category Rules for leather, published in 2018, which assign 88% of farm-level impacts to dairy and 12% to the carcass, then 3.5% of the carcass to hides. That chain implies hides account for 0.42% of the economic value of cattle production. At that share, a hide is close to waste, and buying it changes very little.

Those rules describe European farming, where most cattle are dairy animals. Most of the world's cattle are not, and the best hides come from young beef animals that never produced milk for anyone. Applying a dairy-weighted rule to them is the specific error a 2026 meta-analysis in ACS Sustainable Chemistry & Engineering set out to correct. Using FAO livestock data and regional price data instead, it puts the economic allocation to hides at 8.1% globally, built from reported figures of 6.8% in Italy, 8% in India and 14% in North America.

8.1% against 0.42% is a factor of nineteen. A hide at 0.42% is a scrap the abattoir would otherwise pay to dispose of. A hide at 8.1% is a product with a price, and a price is a reason to keep producing it.

Split by herd, the older figure is narrow rather than wrong. A hide taken from a dairy cow at the end of her milking life is genuinely close to waste: she was kept for milk, and the skin is what remained. That is the case 0.42% describes, and for those hides it holds. It is not where most leather comes from. The trade wants young, unscarred, even-grained hides, and those come from beef cattle who were raised to be killed. The by-product defence describes the exception and gets applied to the rule.

Where durability fits

The cradle-to-gate boundary is where the best argument for animal leather lives. Well-made leather outlasts most PU by years, and a product replaced three times carries roughly three times the footprint. A cheap synthetic bag binned after two seasons is a worse object than one leather bag kept for fifteen.

The arithmetic still does not get there. A durability advantage of three does not close a gap of twelve. It narrows it, and on the older allocation rules it would have closed it comfortably, which is why the two sides of this argument have historically been able to talk past each other while both citing real numbers.

The case against synthetics is unchanged by any of this: fossil-fuel dependence, non-biodegradability, and microplastics shed through wear. Those are real costs that a carbon figure does not capture.

Tanning is dirty on both sides of the ledger. Chrome tanning dominates, and wastewater from one tonne of raw hides carries roughly 240 kg of chemical oxygen demand, 150 kg of suspended solids and around 5 kg of chromate. Treating tannery effluent takes about 3.9 kWh per kg of biodegradable COD removed, against 1.4 kWh for municipal wastewater. Vegetable tanning is the chrome-free alternative and a minority of production, being slower and costlier. It is cleaner rather than clean: the 2026 review notes that vegetable tanning still uses chemical preservatives and finishing agents that stay in the effluent and carry their own ecotoxicity concerns.

What the numbers leave out

Every figure above prices a material. None of them reaches what the material was. A hide is a skin, and it belonged to someone who was bred, confined, transported and killed so that it could be sold. Carbon accounting has no term for that and never will. Buying a hide also settles no account for an animal already dead. A purchase is a price signal, and price signals reach forward, into how many animals are bred next year. That sits outside the allocation dispute, and no result in that dispute answers it.

What should you do?

If your reason is animals: any vegan leather answers the question. Buy the most durable one you can and keep it a long time. Longevity is where synthetics lose, so it is the variable worth optimising.

If your reason is the environment: the label "vegan" is close to uninformative on its own, so ask what the material is. Prefer mycelium or high plant-content options over PVC. Treat "plant-based" as a claim requiring evidence, and assume a PU coating unless the manufacturer says otherwise. On the best current evidence the synthetic still comes out ahead of the hide, but the margin depends on an allocation dispute that is not finished.

If you are buying animal leather anyway: buy it second hand. A used jacket or bag puts no new money into the supply chain and sends no signal back to the farm, and it extends the life of a hide that already exists, which is the outcome every durability argument for leather claims to want. If it has to be new, vegetable-tanned is a materially different product from chrome-tanned and the distinction is usually disclosed, because it is a selling point. But new is where the money reaches the animal, and second hand is where it stops.