Testing protocols and lab routes

We buy what you buy, send it to accredited labs, and publish our methods and raw data. Every study follows the same five-stage protocol, and every named brand gets the right to reply before we publish.

1. Sourcing

• We buy test units anonymously at retail, from at least two retailers or batches where possible.
• We do not use free samples for published results. Samples may be used for pre-screening only.
• We record the retailer, date, price, batch or lot code and photos of each unit and its packaging.

2. Material identification

• FTIR-ATR spectroscopy identifies the polymer (or absence of one) in each component: body, lid, seal, coating, lining, handle, label and packaging.
• Raman spectroscopy is used for very small parts and thin coatings.
• Spectra are matched against reference libraries. We report the match score, and anything below the lab's confidence threshold is reported as "unconfirmed", never as a pass.
• Coatings and linings are sampled separately from the substrate.

3. Release and migration testing (studies only)

• Microplastic release is measured with µFTIR or Py-GC/MS under conditions that mirror home use, with a harsher "worst reasonable case" run reported separately.
• Chemical migration for food-contact items uses the food simulants set out in retained Regulation (EU) No 10/2011.
• Contamination control: glass and metal labware only, filtered water, cotton lab coats, procedural blanks with every batch, and results reported after blank subtraction.

4. Lab routes

We name the lab route in every report. Labs are paid the same fee whatever the result, and never know which brand commissioned a PFH Checked test.

5. Right of reply

1. Before publishing, we send every named brand its results, our method and the exact wording we plan to use.
2. The brand has 10 working days to respond (5 for re-tests). We extend this once, on request, for good reason.
3. We correct any factual error the brand shows us, and we say so in the report.
4. We publish the brand's response in full, or a fair summary with a link to the full text.
5. If a brand does not reply, we say that too.

Brands can submit a response or challenge a finding at any time through our right of reply page.

The research calendar

We run four flagship studies a year — one each quarter — plus smaller product checks each month. PFH Plus members vote on what we test next. See the research calendar.

Publishing data

Every report includes the method, sample size, lab route and limitations. PFH Plus members can download the raw results as CSV. Other researchers can request data for non-commercial use free of charge.