Sampling and test flow in ca proposition 65 compliance: part 2

Key takeaways

1. A test flow has five stages — plan, sample, custody, analysis, verification — and weakness in any one of them undermines the rest.
2. Choose accredited laboratories whose scope covers your methods, and understand how international mutual recognition supports cross-border acceptance of results.^[2]
3. Sampling must cover the lot's full footprint with recorded carton and sample identities, or the report cannot speak for the shipment.
4. Technical regulations operate within WTO disciplines on technical barriers to trade, which is why testing functions as market-access infrastructure.^[1]
5. Verify every report against the plan on lot, component, substance, units, and limits before releasing goods.

Laboratory testing only protects a shipment when the samples genuinely represent the goods and the paperwork proves it. Too many programs treat the lab as the whole of compliance: send something, get a pass, ship everything. The failures that follow are rarely the lab's fault; they come from unrepresentative draws, broken custody, mismatched components, and reports filed without verification.

The full test flow has five stages and the laboratory analysis is only the middle one. Planning defines what will be tested and why. Sampling executes the draw across the lot. Custody carries the samples intact to the bench. Analysis produces numbers. Verification reads those numbers against the plan before anyone releases goods. Weakness in any stage quietly invalidates the others.

This article, the second part of a sampling-and-flow series, walks the whole chain in order. It is written for QC staff and buying offices that want test reports capable of surviving a buyer audit or a marketplace review months after the goods have shipped.

Stage one: write the sampling plan

A one-page plan written before any sample is drawn prevents most downstream arguments. It should name the lot and its carton range, list the components to be tested individually versus as composites, state the substances of concern for each component with a sentence of reasoning, fix the number of samples and their spread across pallets, and describe how retained samples will be sealed and stored. Circulate it to the supplier, the inspector, and the laboratory so all three work from the same page.

Component selection deserves the most thought. Test separately any component with a distinct material or supplier history — coatings apart from substrates, soft plastic parts apart from rigid ones, metal hardware apart from textiles — because compositing across unlike materials dilutes a localized problem into an apparent pass. Where compositing is genuinely appropriate, record exactly which units went into each composite so a failure can still be traced back.

Send the plan to the laboratory for confirmation before shipping samples. The lab checks that the requested methods suit the materials, that detection limits fit the decision at hand, and that reporting units will match any earlier reports for the same product. This single exchange eliminates the most embarrassing class of errors: reports that are technically correct but answer the wrong question. Teams scoping the end-to-end flow can review it at www.utsinspection.com/ca-proposition-65-compliance.html before booking inspection and testing together.

Stage two and three: sample and guard custody

Execute the draw across the lot's full footprint: early, middle, and late cartons, different pallet positions, and coverage proportional to any known production breaks such as shift changes or material batch switches. Number every sample, seal it, photograph the sealed set against the sample list, and log carton numbers alongside. An independent witness at the table converts this routine into evidence, because self-sampling by the factory invites doubts no report can dispel.

Custody between factory and bench needs the same rigor. Log the courier and tracking number, include the plan reference and previous report references in the parcel papers, and ask the laboratory to confirm receipt with sample numbers intact before testing begins. Any discrepancy — a missing seal, a count mismatch — gets resolved before analysis, not rationalized afterward.

Retains close the loop. Keep sealed retained samples from the same draw under controlled storage until the buyer accepts the lot, and reference them in the sampling record. If a result is ever challenged, the retain is what allows re-examination instead of argument, and programs without retains discover this exactly once.

Stage four: choose and brief the laboratory

Laboratory selection rests on two questions: is the lab competent for these methods, and will its results be accepted where they need to be? Accreditation to a recognized laboratory competence standard, assessed by an established accreditation body, answers the first question for the methods inside the lab's stated scope — always check the scope rather than the logo. For the second, the international mutual recognition arrangement for laboratory accreditation provides the framework under which accreditation bodies around the world recognize one another's accredited results, smoothing cross-border acceptance of test data.^[2]

Briefing matters as much as credentials. Tell the lab which product and lot the samples represent, which components and substances are in question, what decision the report will support, and which earlier report reference it continues. Ask for confirmation of method, detection limits, turnaround, and reporting units in writing. A laboratory that understands the decision produces a report shaped for it; one left guessing produces generic numbers that need a second round of explanation.

Keep perspective on why this infrastructure exists. Chemical requirements are technical regulations in the trade-law sense: measures that pursue legitimate objectives such as health protection while disciplining unnecessary obstacles to trade. The WTO agreement on technical barriers to trade sets those disciplines, and conformity assessment — testing, inspection, certification — is the machinery through which exporters demonstrate compliance.^[1] Viewed this way, laboratory fees are market-access costs, and a defensible test flow is the receipt.

How the five stages fit together

Stage five: verify before release

Read the report against the plan line by line: same lot identity, same component descriptions, same substances, same units, and the correct applied limits. Confirm the sample count matches the plan and that any compositing matches what was agreed. If anything diverges, get a written explanation from the laboratory before the lot moves — corrections after shipment carry no weight with an auditor.

Then make the release decision explicit. A short note stating that the lot is released or held, on the basis of which reports, signed and dated, closes the workflow. File the plan, the sampling log, the custody confirmations, and the report as one set keyed to the lot number, so a reviewer months later can reconstruct the entire chain without calling anyone. Shipments protected by files like this survive marketplace reviews and buyer audits; shipments protected by a lone PDF do not.

Frequently asked questions