A harmonized fast-fashion garment-variant dataset for textile circularity and sustainability assessment

Textile circularity research requires product-level data on garment composition and component structure, but most available information is either aggregated at fibre level, derived from physical audits with limited market coverage, or not organized at the colour-variant level needed for sorting and recycling assessment. This data article presents a harmonized fast-fashion garment-variant dataset compiled from publicly accessible Hennes and Mauritz (H&M) and Uniqlo product pages serving the United Kingdom and Australia. Product web addresses (URLs) were collected from 24 to 26 March 2026, and product details were extracted between 24 March and 8 April 2026. Records were filtered, expanded to colour-specific variants where needed, harmonized into a common JSON Lines (JSONL) schema, normalized for material names and garment categories, parsed into component-level material-composition entries, and subjected to consistency checks. After filtering and public-release processing, the dataset contains 47,522 colour-specific garment-variant records. Each record includes source provenance, retailer and region, URL fields and timestamps, gender or section metadata, original category, product name, variant colour, assigned and normalized material-composition text, harmonized parent and detailed garment categories, and structured component-composition entries. The public release excludes product images, raw webpage captures, screenshots, review text, product ratings, review counts, and retailer-specific acquisition scripts. The released files include the final JSONL dataset, processing scripts, mapping tables, processing summaries, citation metadata, license files, and workflow figures. Validation and quality-control information is provided through record-flow summaries, exported mapping tables, assignment-type diagnostics, component percentage-sum flags, and removal counts for unresolved or inconsistent records. The dataset can be reused for sustainability and circularity applications including sorting-compatibility assessment, pre-processing-rule development, design-for-recycling evaluation, fibre-to-fibre recycling analysis, material-flow modelling, life-cycle inventory construction, and comparative studies of online garment information.