DPP for bags and accessories: what changes compared to apparel?

Most writing about the Digital Product Passport assumes you sell garments. A t-shirt has a fibre composition, a country of assembly, a weight and a care label. Feed those into a calculator and you get a footprint number.

A leather tote does not behave that way. It has a hide instead of a fibre, a tanning process instead of a dyeing process, and thirty or so components instead of five. If you make bags, belts, wallets or small leather goods, most steps of DPP preparation differ from the apparel playbook, and a few differ completely.

First, the honest part: nobody has published the rules yet

The Ecodesign for Sustainable Products Regulation (ESPR, Regulation EU 2024/1781) has been in force since 18 July 2024, but it is a framework. The data fields your passport must carry come from a delegated act, a follow-up law written per product group.

The first ESPR working plan, adopted in April 2025, names textiles and apparel as a priority group with an indicative adoption date of 2027. Footwear was left out pending a separate assessment. Bags and leather goods are not named as their own group at all. So when someone quotes you a deadline for handbag passports, ask for the source. As of August 2026 there is no adopted delegated act for textiles, none for footwear, and no published timeline for bags. The reasonable planning assumption is that accessories follow the textile track rather than lead it. That is an assumption, not a date.

What does exist is a signal about content. In May 2026 the Commission’s Joint Research Centre published a study proposing 49 data points in four categories for textile apparel: product identification, producer identification, product information, and compliance documentation. Whatever lands for accessories will most likely be a variation on that list.

The scope split is already visible in customs codes

EU law has already treated leather goods differently from leather apparel in one concrete place. The ESPR ban on destroying unsold consumer products applies from 19 July 2026 for large companies. Annex VII lists the covered codes: HS 4203 (garments and clothing accessories of leather), chapters 61 and 62 (garments), 6504 and 6505 (headgear), and 6401 to 6405 (footwear).

HS 4202 is not on that list. That is the code for trunks, suitcases, handbags and wallets. A leather belt sits inside the destruction ban and a leather handbag does not, even if the same tannery supplied both hides. Your customs classification is doing real regulatory work here, and a catalogue spanning 4202 and 4203 can end up with two compliance profiles under one brand.

Composition: fibre percentages do not translate to leather

For apparel this is a solved problem. Regulation (EU) No 1007/2011 sets the fibre names and requires the familiar percentage breakdown, applying to products that are at least 80% textile fibres by weight.

A structured leather bag rarely clears that threshold, so the textile labelling regime often does not apply in the same way. No EU-wide equivalent forces you to declare “70% full-grain bovine leather, 20% cotton lining, 10% brass hardware” on a label. The one crossover rule runs the other way: where a textile product contains non-textile parts of animal origin, such as a leather patch on jeans, the label must say “Contains non-textile parts of animal origin.” That is a presence flag, not a breakdown.

For a DPP this creates work rather than relief, because the passport asks for material information regardless of what the label regime requires. Expect to define, per style: hide type and animal species, tanning method (chrome, vegetable, chrome-free, combination), finish, lining, thread, and each metal part with its base alloy and plating.

LCA: the leather number is an argument, not a fact

Life cycle assessment (LCA) calculates a product’s environmental impact across its life, from raw material to end of life. For a garment made of known fibres, simplified LCA tools do a decent job.

Leather breaks the tidy version, and the reason is allocation. A cow is not raised to make a handbag. It produces milk, meat, bone and hide, so someone must decide what share of the farm’s emissions belongs to the hide. Under the Product Environmental Footprint category rules used in apparel and footwear, farm-level impacts are split biophysically, with the large majority assigned to milk, and slaughterhouse outputs are then split by economic value.

That last step is where numbers get slippery. A comparison of leather LCA approaches by FootBridge found that moving the economic allocation for hides from 1.8% to 3.5% changed the resulting carbon footprint by roughly 30%. The product did not change. The accounting convention did. So state the method and allocation basis alongside any leather footprint you publish, rather than presenting it as measured fact. Our guide to LCA for fashion DPPs covers how these calculations are structured.

There is also a tooling gap. France’s environmental cost display, set out in Décret n° 2025-957, covers a closed list of eleven garment categories that must be at least 80% textile. Footwear, accessories and leather are excluded, and the public Ecobalyse calculator behind that scheme reflects the same scope.

Hardware, trims and linings: your bill of materials gets long

A jersey t-shirt might have five lines in its bill of materials. A structured shoulder bag can have thirty: outer leather, lining, interlining, reinforcement board, zipper tape, teeth and pull, D-rings, strap clips, feet studs, rivets, magnetic closure, chain, edge paint, adhesive, thread, logo plate. Each comes from a different supplier, often in a different country. Three consequences follow.

Supplier mapping is wider. DPP proposals expect facility-level identification, typically via GLN (Global Location Number, a GS1 identifier for a physical site). A garment might need three or four facility records. A bag can need a dozen, and hardware suppliers are usually the least documented tier in the chain.

Impact modelling is uneven. Public datasets for cattle leather exist. Datasets for a specific zinc alloy buckle with palladium-free plating generally do not, at least not at a resolution matching your part. Some of your bill of materials will be modelled with proxies, and you should record which parts those were.

Changes propagate. Swap a zipper supplier mid-season and you have changed a component carrying its own substance profile and footprint contribution. In apparel that rarely triggers anything. In accessories it can move several passport fields at once.

Substances of concern: a different chemical risk profile

REACH is the EU chemicals regulation. SVHCs are substances of very high concern, listed by ECHA on a candidate list that grows a few substances at a time; the February 2026 update brought it to 253. Check the current list before publishing, because it moves.

Two restrictions matter far more for leather goods than for a cotton shirt.

Chromium VI in leather. REACH Annex XVII entry 47 prohibits placing on the market leather articles, or articles containing leather parts, in skin contact where chromium VI is present at 3 mg/kg or more of total dry weight. It has applied since 1 May 2015 and is tested under EN ISO 17075. Chromium VI is not deliberately added: it can form from the trivalent chromium used in tanning under heat, ageing or the wrong pH. That makes it a batch-variable risk, exactly the kind of parameter a passport is meant to track.

Nickel release from metal parts. Annex XVII entry 27 limits nickel release from articles in direct and prolonged skin contact, including rivets, zippers and fasteners, with a general limit of 0.5 µg/cm² per week measured under EN 1811 (the 2023 version has applied since December 2023). A strap clip or belt buckle sits against skin in ways a garment button often does not.

One detail catches accessory brands out. The 0.1% by weight disclosure threshold under REACH Article 33 applies per article, and following the Court of Justice ruling of 10 September 2015 a component that is an article remains one after assembly. The threshold is assessed against the individual buckle, not the 900 gram bag, so a tiny part can breach 0.1% while the finished product is nowhere near it. Our SVHC and REACH guide covers the disclosure mechanics.

One more piece is still moving. The EU Deforestation Regulation applies from 30 December 2026 for large and medium operators, and cattle products were in scope, which would have pulled hide traceability back to plot-level geolocation. On 13 July 2026 the Commission proposed removing cattle hides, skins and leather from the regulated list. That is a proposal to the Parliament and Council, not a settled outcome, so treat plot-level origin data as valuable but not currently mandatory.

Identification: one product, many components, one GTIN

A GTIN (Global Trade Item Number) is the GS1 barcode number identifying a product. The JRC proposal pairs it with GLN for facilities and HS/TARIC codes for customs. For apparel the logic is well worn: new colour or size, new GTIN. For accessories, two questions recur.

Does a hardware finish change need a new GTIN? If gold and silver hardware are both orderable, they are different trade items to the buyer, so they get different GTINs. Treating hardware finish as a variant inside one GTIN is a shortcut that makes passport data ambiguous later.

What about sets? A wallet-and-cardholder gift set is a predefined assortment in GS1 terms, and the GS1 rules are direct: changing, adding or replacing one item in a predefined assortment requires a new GTIN. The set carries its own GTIN, components keep theirs, and your passport structure has to reflect that nesting. The GTIN and GS1 explainer walks through the numbering.

Batch or unit? Accessories push toward unit

The JRC study recommends production batch as the minimum granularity, with model level for traits that do not vary across a run and item level staying voluntary.

For a 20,000 unit t-shirt order, batch is obviously right. For accessories, three things pull the other way. Volumes are lower and values higher, so a unique code per unit is a smaller share of the item price. Hides genuinely vary, since two bags from the same order can come from different hide lots and tanning batches. And resale matters more for bags than for most garments, where unit-level identity helps with authentication and ownership history.

None of that is required. Going finer than the floor is a business decision, and it is reversible in one direction only, because you cannot retrofit unit identity onto stock already shipped.

How Wetrack handles this

Wetrack is a DPP platform for fashion brands, taking you from product import to published, EU-compliant passports with QR codes. On the specific gaps above:

• Leather and bags LCA. Alongside garment LCA calculated via Ecobalyse (CO2, PEF and durability), Wetrack ships a dedicated leather LCA provider and a bags LCA, so leather goods are not left without a calculation path when the general textile calculator excludes them.
• Granularity you choose. Custom QR codes are generated as SVG at product, batch or unit level, so the batch-or-unit decision is a setting rather than a rebuild.
• Supplier records. The shared supplier database is built on Open Supply Hub, which helps when a bag pulls in more facilities than a garment.
• Claims discipline. Sustainability claims require certification files. Without one, a claim is labelled “Self-Declared” rather than presented as verified.
• No lock-in. Data follows the ODSAS standard and a REST API is available.

Wetrack does not run chemical testing, issue certificates or handle EUDR due diligence. Those stay with your labs and suppliers. There is a live sample passport if you want to see the output first.

FAQ

Are bags and accessories covered by the EU Digital Product Passport? Not yet by any adopted delegated act. ESPR names textiles and apparel as a priority group with an indicative 2027 adoption date, footwear is under separate assessment, and bags have no published timeline of their own. The working assumption is that accessories follow the textile requirements rather than get a separate regime, but that is not confirmed.

Do I need a different LCA method for leather products? Practically, yes. Leather LCA depends on allocation choices for hides that do not arise for textile fibres, and those choices can shift the footprint materially. The general calculators used for garments, including the public Ecobalyse tool, currently exclude leather and accessories from their scope.

What substance restrictions apply to leather bags specifically? The two most relevant are chromium VI in leather in skin contact, restricted to below 3 mg/kg of dry weight under REACH Annex XVII entry 47 since May 2015, and nickel release from metal parts in prolonged skin contact, generally limited to 0.5 µg/cm² per week under entry 27. The SVHC candidate list also applies, assessed per component rather than per finished bag.

Should my accessory passports be batch level or unit level? Batch is the proposed minimum. Unit level is worth considering because volumes are lower, hide variation is real, and resale benefits from a unique identity. It costs more per item and cannot be applied retroactively, so decide before your next production run.