Definition · Plain-language
Out-of-specification (OOS)
An out-of-specification result is a test result that falls outside the approved limits or acceptance criteria, requiring a formal investigation.
The step most authors miss
Doing CRediT right? Don’t stop at the statement.
A CRediT statement credits you inside one paper. The recognition CRediT was built for happens when those roles are tied to you, persistently. Sign in with your ORCID — free — and claim your CRediT contributions on casrai.org, the home of the standard. They become a verified, portable part of your identity, not a line that disappears into one PDF.
Free: claim your contributions, then export a journal-ready CRediT statement, schema.org structured data, JATS XML, CSV or BibTeX — and preview your public profile. A membership publishes that profile publicly and verifies the journals you serve.
What counts as OOS
A result is out-of-specification when it lies outside the predetermined specifications or acceptance criteria set for the material, in-process check, finished product or stability sample. The specification is the standard the result is judged against; crossing it is the trigger. Related terms describe nearby situations: out-of-trend (OOT) results sit within specification but deviate from the expected pattern, and out-of-expectation results are atypical. OOS is the formal category that obliges the firm to investigate before deciding what the result means.
The investigation
A widely followed framework comes from FDA guidance on investigating OOS results. The first phase is a laboratory investigation to determine whether an assignable laboratory error — such as a dilution mistake or instrument fault — caused the result, before any retesting. If no clear laboratory cause is found, the investigation expands to a full review covering production, the batch record and other batches, with retesting and resampling conducted only under a predefined, scientifically justified plan. The conclusion, and any decision to invalidate the original result, must be documented and approved.
Why discipline matters
OOS handling is a sensitive area for data integrity. Discarding an unfavourable result, retesting until a passing value appears, or invalidating results without a sound investigation are serious breaches that regulators actively look for. Every original result, including the OOS one, must be retained and accounted for. A robust OOS procedure links to deviation management, root cause analysis and CAPA, so that genuine problems are corrected and the eventual disposition of the batch is defensible.
Key facts
At a glance
- Definition: test result outside approved specifications
- Abbreviation: OOS
- Trigger: a documented investigation under GMP
- Related: out-of-trend (OOT), out-of-expectation results
- Framework: FDA OOS investigation guidance (lab then full review)
- Caution: no retesting away or invalidating without justification
Common misconceptions
What people often get wrong
Often heard: An out-of-specification result can be retested until it passes.
Actually: Retesting is allowed only under a predefined, justified plan after investigation; testing into compliance to discard an OOS result is a serious data-integrity breach.
Often heard: OOS just means the product has failed and the batch is rejected.
Actually: An OOS result triggers an investigation to determine whether it reflects a true product problem or a laboratory error before any disposition decision is made.
Often heard: Out-of-specification and out-of-trend mean the same thing.
Actually: OOS results fall outside specifications; out-of-trend (OOT) results remain within specification but deviate from the expected pattern.
Going deeper
