Proficiency Testing and Interlaboratory Comparisons: What Supplement Brands Should Be Asking Their Lab

Every year, hundreds of dietary supplement brands submit their products to a testing lab and receive a Certificate of Analysis declaring everything within spec. And every year, a subset of those same products fail when tested independently — by an FDA investigator, a retailer’s quality team, or a second lab contracted during a dispute.

The gap between “our COA says it’s fine” and “it actually is fine” is where proficiency testing programs live. It’s a gap that far too many supplement brands never think to ask about — and one that a lab’s accreditation certificate alone will never close.

What Proficiency Testing Is — And Why ISO 17025 Requires It

Proficiency testing (PT) is the practice of sending blind, homogeneous samples to multiple labs simultaneously, then comparing results to see which ones land within an acceptable range. The labs don’t know who else got the sample. They don’t know what the “correct” answer is ahead of time. They run the analysis and report back.

The statistical engine behind most PT schemes is the z-score. A z-score between -2 and +2 is generally considered satisfactory — meaning your result falls within two standard deviations of the assigned value. A score outside that range is flagged as unsatisfactory, which triggers a mandatory investigation and corrective action.

That requirement isn’t buried or discretionary. Section 7.7.1 of ISO/IEC 17025:2017 explicitly states that labs shall “monitor their performance by comparison with results of other laboratories” through participation in PT or interlaboratory comparisons (ILC). For labs seeking or maintaining accreditation through organizations like A2LA (American Association for Laboratory Accreditation) or Perry Johnson Laboratory Accreditation (PJLA) — both active in California and across the US — PT participation is a scored element of the assessment. A lab that can’t produce a history of PT participation for a given method should give you pause.

The leading PT providers in the food, dietary supplement, and cosmetic testing space include:

• AOAC INTERNATIONAL — offers schemes covering vitamins, minerals, and mycotoxins
• FAPAS (operated by FERA Science) — widely used for food chemistry, pesticide residues, and heavy metals
• NSF International — runs PT programs specifically targeting nutraceutical and supplement matrices
• AAFCO — covers feed and agriculture-adjacent analytes

Most accredited labs participate in at least 2–3 PT rounds per analyte category per year. For high-volume labs running hundreds of samples weekly, that number climbs considerably — and so does the depth of benchmarking data available to show you.

Why PT Results Tell You Things a Lab’s Accreditation Certificate Doesn’t

An accreditation certificate from A2LA or PJLA tells you that a lab has a documented quality system, trained personnel, and appropriate equipment. It does not tell you how that lab performs on the specific analyte in the specific matrix relevant to your product today.

That’s what PT data tells you.

Consider vitamin D3 testing in softgel matrices — one of the more technically demanding assays in the supplement space, given the lipid-heavy matrix and the susceptibility of cholecalciferol to oxidation during sample prep. Labs that don’t routinely run softgel matrices will often see elevated extraction variability, which shows up as z-score drift in PT rounds. The accreditation certificate doesn’t capture that. The PT history does.

The same story plays out with heavy metals in protein powders. ICP-MS is the gold standard instrument for measuring arsenic, lead, cadmium, and mercury at parts-per-billion levels, but calibration curve construction, matrix-matched standards, and proper use of internal standards like bismuth or indium all affect accuracy in ways that aren’t visible from a scope-of-accreditation document. A lab running 500 samples per week with rigorous QC checkpoints looks very different in a PT round than one running the same volume with minimal quality oversight — even if both hold identical accreditation status.

We’ve seen labs with perfectly valid ISO 17025 certificates produce z-scores above +3 for arsenic in botanical matrices. That’s more than three standard deviations from the reference value — in practical terms, a result far enough off to matter for Prop 65 compliance, FDA review, or retailer qualification. A supplement brand relying exclusively on the certificate to assess lab quality would have no way to detect that.

The Questions to Ask Your Lab Before Signing a Testing Agreement

If you’re evaluating a contract lab for supplement testing, regulatory compliance, or Amazon Seller Central qualification, these are the PT-specific questions worth asking before you commit:

1. “For the analytes I need tested, what PT programs do you participate in — and can you share your z-score history for the past 12 months?”

A reputable lab will have this data organized and should share it without hesitation. Redacting client-identifying information is reasonable; a flat refusal is a red flag. Look for consistent satisfactory ratings across rounds, not just a single clean result.

2. “Have you received any unsatisfactory PT results in the past year for methods relevant to my product category?”

Unsatisfactory results aren’t automatically disqualifying. What matters is how the lab responds — whether they document the nonconformity, investigate the root cause, and implement a corrective action. That process is exactly what ISO 17025:2017 Section 8.7 requires, and a lab that can walk you through it is demonstrating precisely the quality culture you want behind your testing data.

3. “What’s your PT coverage for biological contaminant assays, specifically for the matrices I’ll be testing?”

Microbial PT — covering Salmonella, E. coli, Staphylococcus aureus, and yeast/mold counts — is often less robust at smaller labs than chemical PT. If you’re testing finished dietary supplements that need USP <61>/<62> compliance, you want a lab that’s routinely benchmarking against other qualified labs on those exact methods, not just on chemistry assays.

If the lab struggles to answer any of these questions fluently, that tells you something. It doesn’t necessarily mean they produce bad data. But it does mean they haven’t built the continuous quality monitoring infrastructure that high-stakes supplement testing demands.

The Corrective Action Loop: What Happens When a Lab Gets a Bad PT Result

This is the part most brands never see — and arguably the most important indicator of a lab’s quality culture.

When a lab receives an unsatisfactory PT result, they’re required under ISO 17025:2017 Section 8.7 to open a nonconformity record, investigate the root cause, and document corrective actions before the issue is closed. Common root causes we encounter in interlaboratory comparisons include:

• Calibration curve errors or insufficient calibration ranges for the concentration levels expected in the matrix
• Matrix-specific extraction losses not accounted for in recovery studies
• Reference material purity assumptions that turn out to be inaccurate at the batch level
• Instrument drift occurring between scheduled calibration events

The corrective action process isn’t just about fixing the specific problem. It requires asking whether the same root cause could have affected client samples during the relevant period. Most A2LA-accredited labs have a documented client notification policy covering exactly this scenario — and that policy should exist whether you ask about it or not.

From a regulatory standpoint, this matters. FDA 21 CFR Part 111 (dietary supplement CGMP regulations) requires manufacturers to use “qualified laboratories” for testing. While the regulation doesn’t explicitly define qualification in terms of PT participation, it’s increasingly clear through FDA inspection observations and third-party quality audits that PT history and corrective action records are viewed as evidence of qualification. We’ve had quality auditors request PT records during CGMP supplier audits, and labs that can’t produce them face uncomfortable scrutiny that eventually lands on the brand that hired them.

What Interlaboratory Comparisons Add That Traditional PT Doesn’t

PT programs use certified reference materials and statistically assigned values, which makes them excellent for calibration benchmarking. But they don’t always replicate the complexity of real product matrices. Interlaboratory comparisons (ILCs) fill that gap.

In an ILC, a set of labs test the same actual product sample — often a real commercial product rather than a prepared reference material — and compare results descriptively rather than against a fixed correct answer. This makes ILC data more relevant for understanding between-lab variability on complex matrices like botanical extracts, emulsified topicals, or protein powder blends.

A single analyte — total aerobic plate count in a powdered greens supplement, for example — can show 10-fold variation between labs depending on growth media selection, incubation temperature, and dilution scheme. ILC data quantifies that variability in a way accreditation documentation can’t. For high-risk or high-volume product categories, some brands now require their labs to participate in ILCs as a contractual condition — particularly when those products are subject to FDA scrutiny or sold through retailers with independent testing programs.

It’s also worth knowing that some ILC programs are category-specific. If you’re selling a probiotic, a protein blend, and a herbal extract, the matrices are different enough that ILC performance in one category doesn’t automatically transfer to another. Ask specifically about the product types your lab has participated in ILCs for.

What to Actually Do With This Information

You can’t audit every lab that handles your product. But you can ask better questions before you sign a testing agreement — and you can treat the answers as real data points rather than formalities.

Request PT participation records for the specific methods you care about. Ask about the lab’s nonconformity and corrective action process for unsatisfactory rounds. Find out whether they use AOAC, FAPAS, or NSF PT programs — or some combination. Ask whether they’ve participated in any ILCs for matrices similar to yours. And if the lab struggles to answer those questions, treat that difficulty as the quality signal it is.

A Certificate of Analysis is only as trustworthy as the data behind it. Proficiency testing is how labs prove — to themselves and to you — that their data is worth trusting.

---

Written by Nour Abochama, Vice President of Operations, Qalitex Laboratories. Learn more about our team

Talk to our team about your testing needs. Contact us

Related from our network

• FDA Audit Readiness and Quality System Documentation for Supplement Brands — Aurora TIC covers the regulatory consulting side of FDA inspections, CAPA programs, and quality system design.
• Supplier Qualification and COA Verification for Raw Material Procurement — Ayah Labs explores how incoming raw material testing and supplier audits should be structured before product manufacturing begins.
• Canadian GMP and Health Canada Compliance for NHP Manufacturers — Androxa breaks down what Health Canada’s Natural Health Products Directorate expects from testing documentation and licensed facilities.