When Your Supplement Test Comes Back Failing: Inside the OOS Investigation Process

A supplement brand ships samples to a testing lab on a Tuesday. By Thursday, they have a certificate of analysis showing a total aerobic count of 1,800 CFU/g — well above the USP <61> acceptance criterion of 1,000 CFU/g for oral non-sterile powders. Their contract manufacturer swears the batch is clean. The lab says the result stands.

Now what?

This scenario lands in our queue more often than most brands expect. And the uncomfortable reality is that how a laboratory handles a failing result — what happens in the hours after that COA is generated — tells you far more about their competence than any accreditation certificate on the wall. The COA is the artifact. The investigation file is the evidence.

What “Out of Specification” Actually Means in a Supplement Testing Lab

An out-of-specification (OOS) result is any test result that falls outside the acceptance criteria established in a product specification, a pharmacopeial monograph, or a regulatory limit. In dietary supplement testing, that reference point might be a USP <61> microbial limit, a Prop 65 threshold for lead in a botanical, an in-house specification for potency, or an identity standard required under 21 CFR Part 111.

What OOS does not automatically mean is that the product is defective.

That distinction matters enormously — for brands deciding what to do next and for FDA compliance purposes. Roughly 30–40% of initial OOS results in analytical chemistry can be traced back to laboratory error: a miscalibrated instrument, a sample preparation mistake, an analyst technique issue, or something as mundane as a mislabeled reagent. In microbiology, preservative-rich matrices like certain gel caps or oil-based softgels can interfere with growth media, artificially elevating counts. A competent lab knows this, and their first obligation is to investigate themselves before drawing conclusions about your product.

The formal framework for this comes from FDA’s 2006 Guidance for Industry on Investigating Out-of-Specification Test Results for Pharmaceutical Production. Written for drug manufacturers, it’s widely applied by analogy in the supplement and food testing space. ISO 17025:2017 — the international accreditation standard governing testing laboratories — separately requires documented procedures for handling nonconforming work, which includes unexpected or failing results. Both frameworks point to the same logic: structure the investigation, document everything, don’t jump to retesting as a first response.

Phase 1: The Laboratory Investigation — Where Every OOS Should Start

Phase 1 is entirely internal to the lab. No retesting a new sample. No calls to the client yet. The focus is deliberately narrow: did the laboratory make an error?

A properly structured Phase 1 investigation covers six areas:

Analyst review. Was the correct procedure followed step by step? Were instrument parameters set correctly? Did the analyst note anything unusual in the run log — a pressure spike, a baseline drift, a peak that looked off?

Instrument verification. Were calibration checks and system suitability criteria passing at the time of analysis? Are there trending issues in the instrument’s recent performance records? A single point calibration drift might not trigger an alarm on its own but could be the explanation when viewed in context.

Sample preparation review. Was the correct amount weighed? Were reagents in specification and within their use-by date? Did the sample look unusual during prep — unexpected clumping, discoloration, phase separation?

Reference standard check. For potency analyses, was the working standard at the correct concentration? Properly stored? Has the lot been used without incident on other recent analyses?

Calculation audit. Modern LIMS-integrated labs catch most arithmetic errors automatically, but manual transcription between instruments and systems still happens. Every formula and data transfer should be verified.

Assignable cause determination. If any of the above reveals a clear, documentable error — something specific, not “we think maybe” — the original result can be invalidated and the analysis repeated with proper documentation.

The hard rule here: Phase 1 cannot conclude there’s no error just because investigators didn’t find one. Absence of evidence isn’t evidence of absence. If no assignable cause is identified, the result stands and Phase 2 begins.

At an ISO 17025 accredited supplement testing lab, every step of this process must be documented in a nonconforming work record — not a sticky note, not a verbal exchange, but a traceable written record that an assessor could pull and audit two years later.

Phase 2: The Full-Scale Investigation — When Phase 1 Doesn’t Resolve It

Phase 2 is where things get collaborative and sometimes uncomfortable. At this stage, the failing result is presumed real, and the investigation expands beyond the lab’s walls.

This is when we call the brand and start asking pointed questions: What does the batch record show? Was there a deviation during manufacturing? How was the sample collected from the lot — who collected it, and were they trained on sampling procedures? How was the sample packaged and shipped to the lab — was it subject to heat, extended transit, or inadequate cold chain?

In pharmaceutical manufacturing, Phase 2 investigations are highly formalized, with 30-day completion timelines enforced by quality units. Supplement brands operating under 21 CFR Part 111 have somewhat less prescriptive timelines, but the underlying standard is the same: every confirmed OOS result should have a documented investigation with a root cause or a documented explanation of why the root cause couldn’t be determined.

One pattern we see brands fall into: assuming that a retest from a different portion of the same lot automatically resolves the issue. It doesn’t — at least not the way most people think. Averaging a failing result with passing retests to produce a net-passing average is not scientifically valid and not regulatorily acceptable. FDA investigators look specifically for this practice when reviewing OOS records during inspections.

Additional testing drawn from multiple sample units across the lot can be appropriate, but only when there’s statistical justification, a defined sampling plan, and documented rationale established before any additional testing begins. An after-the-fact retesting strategy constructed to reverse a failing result is exactly what a 483 observation looks like.

Documentation Requirements Under 21 CFR Part 111

Under 21 CFR Part 111 Subpart J, dietary supplement manufacturers — and by extension their contract testing labs — must establish and follow written procedures for laboratory operations, including how OOS results are handled. Sections 111.303 and 111.325 require that all test results be reviewed before a batch is released, and that any result outside specification triggers an investigation.

What FDA expects to find when they pull your records:

• A written OOS SOP with defined phase structures and decision points, signed and dated
• Completed investigation forms for every OOS event, with analyst name, supervisor review, and disposition decision
• A conclusion that is either “laboratory error — result invalidated” with documented, specific evidence, or “confirmed OOS” with a corresponding batch disposition record
• Retention of investigation files for at least one year beyond the product’s shelf life, or two years from date of distribution — whichever is longer

In fiscal year 2024, FDA issued 74 warning letters to dietary supplement companies. Inadequate laboratory controls — including failures in OOS investigation — ranked consistently among the most-cited violations, alongside labeling deficiencies and cGMP documentation gaps.

The specific language you do not want appearing in a warning letter: “Your firm failed to thoroughly investigate any unexplained discrepancy.” It’s concise. It’s damaging. And it almost always follows an inspection where OOS records either didn’t exist or didn’t hold up to scrutiny.

Five Questions to Ask Your Testing Lab Before the Next OOS Happens

Most supplement brands never think to ask their lab about OOS procedures — until they’re staring at a failing COA on a Friday afternoon with a shipment scheduled Monday. These are the questions worth asking upfront, before you need them.

Do you have a written OOS SOP you can share? Any ISO 17025 accredited supplement testing lab should produce this without hesitation. If there’s resistance — claims that it’s proprietary, or vague assurances that “we follow industry standards” — that’s worth weighing carefully.

How do you distinguish between Phase 1 and Phase 2 investigations? A lab that treats all OOS events identically, or defaults immediately to retesting, probably isn’t running structured investigations. Ask for an example of how a recent investigation was documented.

Who reviews and approves OOS investigations before the final report is issued? It should be a quality manager or technical lead who was not the analyst of record on the original test. Separation of duties in the review process is a basic quality control principle.

How long do you retain OOS investigation records? The answer should be at least as long as the regulatory retention requirement for the product — not “30 days” or “as long as the client needs them.”

What happens to my retained sample if a result comes back OOS? Many labs dispose of samples on a routine schedule regardless of result status. For any OOS investigation, the sample should be retained until the investigation is formally closed — and you should know when that is and who makes that call.

The labs that can answer all five questions clearly, with specific references to their own SOPs, are the ones worth building long-term relationships with. The ones that fumble or pivot to turnaround time and price are telling you something, too.

---

If you’re working through an OOS result and aren’t sure whether your lab’s investigation meets the standard — or if you’re evaluating a new testing partner and want to understand what rigorous OOS handling looks like in practice — we’re glad to walk through our procedures with you.

Talk to our team about your testing needs. Contact us

---

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

Related from our network

• Building Audit-Ready Documentation for FDA Inspections — Aurora TIC covers the regulatory consulting side of 21 CFR compliance, including how to design investigation SOPs that hold up under FDA scrutiny.
• Raw Material COA Verification and Supplier Qualification Testing — Ayah Labs specializes in ingredient-level certificate of analysis verification and supplier qualification — the upstream quality layer that reduces OOS events before manufacturing begins.