Open data catalog
Download Laboratory.love results data
Download the public results as CSV or JSON. Both files contain the same measurement-level records: one row or object for each product result, physical sample, and analyte.
Version 2026-09-12: 8,456 measurements across 248 results — 12 commissioned by Laboratory.love and 236 derived from PlasticList.
Dataset downloads
Download CSVTabular long-format data for spreadsheets and analysis tools.Download JSONDataset metadata and the same records for software and agents.What the files contain
Identity
result_id, result_url, product_name, brand, sample_id, and lot_identifier.
Provenance
source_type, source_name, source_url, license_url, and attribution.
Testing context
purchase_date, test_date, assay, method, and laboratory.
Measurement
analyte, abbreviation, category, value, qualifier, status, LOD, LOQ, and unit.
Concentrations use ng/g, numerically equivalent to parts per billion for this purpose. Blank values are intentional: the files do not infer missing dates, methods, laboratories, detection limits, or measurements. Use the status and qualifier fields when interpreting a value.
Provenance and license
Laboratory.love-commissioned results and the original dataset compilation are available under Creative Commons Attribution 4.0 International. Reuse instructions: credit "Laboratory.love", link to the relevant Laboratory.love result or dataset page, link to the CC BY 4.0 license, and indicate whether changes were made.
PlasticList-derived records retain their own PlasticList attribution and source URL. Per-record source and license fields make the boundary explicit. Factual measurements may not themselves be copyrightable; the license applies to the extent copyright or database rights exist.
Preferred citation
Laboratory.love. “Plastic-Chemical Test Results,” version 2026-09-12. https://www.laboratory.love/data
Version and interpretation
Current version: 2026-09-12; last modified 2026-09-12. Files are regenerated from the same normalized result records used to build the public result pages, preventing a separate hand-maintained dataset from drifting.
See the methodology for sampling, assays, non-detect handling, exposure calculations, limitations, and correction history. These results are point-in-time measurements of specific samples, not a guarantee about every unit of a product.