As Found / As Left in Calibration: 5 Critical Facts That Protect Your Quality Program

As Found / As Left in Calibration are the two most important entries on any calibration certificate — yet most teams overlook what they actually reveal.
. It reached USD 5.7 billion in 2023 and is projected to hit USD 8.1 billion by 2030 — a compound annual growth rate driven by tightening product quality standards in aerospace, medical devices, and precision manufacturing (MarketsandMarkets, 2023). Behind that growth sits a question every quality manager must answer: how do I know my instruments were accurate before calibration — not just after?

That’s exactly what as-found and as-left calibration data answer. These two entries on a calibration certificate are the most information-dense fields in the document, yet many organizations treat them as a formality. They aren’t. They are the difference between knowing your measurement process was trustworthy and merely hoping it was.

This guide explains what each term means in plain language, why every calibration certificate must contain both readings, how to interpret them correctly, and what your team should do the moment an as-found reading reveals a problem.

In 2026, ISO 10012:2026 defines as-found data as the measurement result recorded for an instrument before any calibration, adjustment, or repair takes place — a pre-service snapshot compared against a NIST-traceable reference standard to determine actual error at the time of receipt. This reading reveals the true condition of the instrument while it was actively performing measurements in your facility.

The purpose is straightforward: before the technician touches anything, measure the instrument’s current performance. Compare that performance to the specified tolerance. Record what you find — exactly as you find it.

Example A tells a reassuring story: the wrench was −2.8% off the reference load — well within its ±4% tolerance — while it was being used on the production floor. Example B is a different matter entirely. An −8.2% error on a joint torqued to specification means every fastener tightened by that wrench during the last calibration period may have been under-torqued. That’s not a calibration issue. That’s a potential product safety issue.

As-found data matters beyond the single measurement event. Over multiple calibration cycles, the pattern of as-found readings reveals whether an instrument is drifting steadily toward the tolerance limit (a sign the interval should be shortened), or remaining stable through its interval (a sign the interval may safely be extended). No other data source provides this insight as directly. For a deeper look at how as-found patterns factor into interval decisions, see Micro Precision’s Calibration Interval Determination: Risk-Based Guide.

As-left data records the instrument’s measurement performance after all calibration work is complete — after adjustment, repair, or confirmation that no adjustment was needed. Where as-found is the opening diagnostic, as-left is the closing certificate: it documents that the instrument has been brought into conformance and is fit to return to service.

If an instrument arrives within tolerance and no adjustment is needed, the as-found and as-left readings will be identical — or within measurement uncertainty of each other. This is a good outcome. It means the instrument remained stable through its entire calibration interval, and that data point contributes positively to its reliability history. Don’t interpret identical readings as a sign the calibration was unnecessary; they confirm the previous interval was correctly set.

The as-left record answers the auditor’s most fundamental question: Is this instrument currently accurate enough to trust? Without it, a calibration certificate is incomplete regardless of what standards or specifications it references on the cover page. Understanding the distinction between accuracy, precision, and resolution in calibration helps clarify exactly what the as-left reading is — and isn’t — claiming.

A calibration certificate that reports only the post-calibration (as-left) condition tells you the instrument is accurate now — but cannot tell you whether it was accurate during the weeks or months it was in use. That gap isn’t academic. It’s a regulatory exposure and a product quality risk that auditors are specifically trained to identify.

Consider two instruments. Instrument A shows a 0.3% error as-found and a 0.1% error as-left — stable, in-tolerance, clean history. Instrument B shows a 7.8% error as-found (out of tolerance) and a 0.1% error as-left after adjustment. Looking only at as-left, both certificates appear identical: both instruments leave the lab accurate. With as-found data, the story is complete — and Instrument B triggers an investigation.

In 2026, ISO 10012:2026 — the second edition of the international standard for measurement management systems, replacing the 2003 version — was published with explicit requirements for data-driven calibration interval validation. The new edition aligns with the harmonized management system structure of ISO 9001 and requires organizations to use historical calibration data, including as-found results, to confirm that calibration intervals are still appropriate (ANSI Blog, ISO 10012:2026, 2026). Without as-found data on every certificate, organizations cannot satisfy this requirement — because the data doesn’t exist to analyze.

When an instrument’s as-found reading falls outside its specified tolerance, the calibration finding triggers a structured response that extends well beyond the instrument itself. In 2026, more than 1.3 million organizations worldwide certified to ISO 9001 must maintain documented procedures for exactly this situation (ISO Survey, simpleQuE, 2024).

The response has five defined stages. Each must be documented to satisfy ISO 9001, FDA QMSR, and most aerospace quality requirements.

Remove the instrument from service immediately. Do not allow measurements taken after the last confirmed in-tolerance calibration to pass through to a finished product, released report, or shipped assembly without formal review.

Determine the date of the instrument’s last known in-tolerance calibration — the as-left date from the previous cycle. Identify every product, batch, assembly, or measurement record created using this instrument between that date and today.

Assess whether the measured error was large enough to affect product conformance. A torque wrench −3% out of tolerance on a joint with a ±10% process tolerance likely did not create nonconforming product. A pressure gauge −15% out of tolerance on a critical structural joint is a different calculation entirely.

Document the finding, the impact assessment outcome, root cause investigation, and corrective action. This NCR becomes part of the instrument’s permanent calibration history and may be requested by customers, auditors, or regulators.

Use the out-of-tolerance as-found finding as input to shorten the instrument’s calibration interval per NCSL RP-1 methodology. The interval that allowed drift this far is demonstrably too long for this instrument in this environment.

ISO/IEC 17025:2017 requires accredited calibration laboratories to include measurement results on every calibration certificate — and as-found/as-left data are the most direct way to fulfill that requirement. As of March 2025, more than 114,600 laboratories worldwide hold ILAC MRA accreditation under this standard, spanning every industry that depends on measurement integrity (ILAC, ISO, March 2025).

Here’s how each major standard applies:

All as-found and as-left measurements only carry legal and technical weight if the reference standards used are traceable to the National Institute of Standards and Technology — or an equivalent national metrology institute — through an unbroken chain of calibrations. NIST’s traceability policy defines this chain explicitly (NIST, Traceability Policy). Without traceability, as-found and as-left data, no matter how carefully collected, cannot be used to make defensible quality decisions or satisfy regulatory requirements. The numbers are only as meaningful as the standards that produced them — see our breakdown of primary, secondary, and tertiary calibration standards to understand exactly how that traceability chain is built.

As-found and as-left calibration data aren’t administrative formalities — they’re the evidentiary backbone of a defensible quality program. As-found tells you whether your instruments were trustworthy while they were doing actual work. As-left confirms they’re trustworthy when they leave the lab. Together, they give you what no other single document can: a complete, auditable measurement accuracy history.

With ISO 10012:2026 now in effect and the FDA QMSR operative since February 2026, the documentation stakes are higher than at any point in the last decade. Organizations that treat calibration records as a regulatory checkbox are accepting risk they can’t see — and paying for investigations they could have prevented.

The global calibration services market reflects this growing rigor, projected to reach $8.1 billion by 2030 as aerospace, medical device, and precision manufacturing industries double down on measurement integrity. Your as-found and as-left records are how you prove — to auditors, to customers, and to yourself — that your measurement systems are actually doing their job. If you want to go deeper on what drives instruments out of tolerance between calibration cycles, our article on calibration drift — its causes, detection, and prevention covers exactly that.