Compliance and audits
How to set calibration intervals
No standard tells you to calibrate every twelve months. They require calibration at defined intervals and expect the interval to have a documented, defensible basis. This guide explains how to set that interval, refine it with your own data, and defend it at audit.
Why there is no universal interval
A calibration interval is how often an instrument is recalibrated. No quality standard sets a fixed number. ISO 9001, ISO/IEC 17025, ISO 13485, AS9100, and IATF 16949 all require calibration at defined intervals and expect the interval to rest on a documented basis rather than a habit. A flat one-year default is common, but it is weak to justify and often either wastes money on stable instruments or lets critical ones drift too long.
The right interval balances the cost and downtime of calibrating against the risk of using an instrument that has drifted out of tolerance since its last calibration.
The inputs that set a defensible interval
Whatever method you use, these are the factors that justify an interval. Record which ones drove your decision so the interval holds up when an assessor asks.
- Manufacturer guidance. The maker's recommended interval is a reasonable starting point, especially before you have history.
- How critical the measurement is. Tighter tolerances and higher product or safety risk argue for shorter intervals.
- How the instrument is used. Heavy use, harsh environments, transport, and shop-floor handling all shorten the interval an instrument can safely hold.
- Its own history. How often the instrument has come back in or out of tolerance at calibration is the strongest evidence of how long it can hold its interval.
- Regulatory or customer requirements. A contract or regulation may fix a maximum interval regardless of your analysis.
Method 1: Start from guidance, then refine
When an instrument is new and you have no history, start from the manufacturer's recommended interval plus any customer or regulatory requirement, then adjust as data accumulates. Recognized interval-analysis frameworks such as NCSLI RP-1 and ILAC-G24 describe this approach. It is defensible on day one and gets better as you gather results.
Method 2: A reliability-based interval from your own data
Once you have several calibrations for an instrument, set the interval so a target in-tolerance reliability is maintained, for example keeping at least 90 to 95 percent of calibrations in tolerance as-found. If too many come back out of tolerance, the interval is too long and you shorten it. If an instrument is reliably in tolerance calibration after calibration, the interval may be safely lengthened, saving cost and downtime.
This is the most defensible method because the number comes from the instrument's own behavior. Axiospec shows each instrument's as-found in-tolerance rate on the asset, and once something has come back out of tolerance it computes a reliability-based recommended interval against a 90 percent in-tolerance target, so the evidence sits next to the asset. The interval you set stays your decision.
Watch drift, not just pass or fail
An instrument that passes every calibration but drifts steadily toward its tolerance limit is telling you the interval is getting risky before it ever fails. Charting the as-found reading against the tolerance band over time surfaces that trend early, so you can shorten the interval before an out-of-tolerance event affects product.
Adjusting intervals up and down
Intervals are not set once. Lengthen them when history clearly supports it, to cut cost and downtime. Shorten them after an out-of-tolerance result, when an instrument is used harder, or when the measurement becomes more critical. The important part for an audit is that each change has a recorded basis, so the interval is never an unexplained number.
What auditors look for
An assessor is not checking for a specific interval. They are checking that you have a documented, consistently applied basis for the intervals you set, that you review them against each instrument's history rather than leaving them fixed forever, and that you act when equipment is found out of tolerance. A calibration record that carries the interval, the basis, and the as-found and as-left readings answers all three.
How Axiospec helps
Axiospec shows the as-found in-tolerance rate and, once an instrument has come back out of tolerance, computes a reliability-based recommended interval and charts drift from each instrument's own calibration history, sets the next due date automatically from the interval and last calibration, and surfaces what is due, due soon, and overdue on a Due Calendar. Every interval and the basis for it lives on a tamper-evident record. The analysis is advisory, so the call stays yours.
Common questions
- How do you choose a calibration interval?
- Start from the manufacturer's recommended interval plus any customer or regulatory requirement, then refine it with your own calibration history. The factors that justify an interval are manufacturer guidance, how critical the measurement is, how hard the instrument is used, its own in-tolerance history, and any contractual or regulatory maximum. Record which of these drove the decision. An assessor is not checking for a specific number. They are checking that you have a documented, consistently applied basis for it.
- Is a calibration interval required to be 12 months?
- No. No quality standard sets a fixed number. ISO 9001, ISO/IEC 17025, ISO 13485, AS9100, and IATF 16949 all require calibration at defined intervals and expect the interval to rest on a documented basis rather than a habit. A flat one-year default is common but weak to justify, and it often either wastes money on stable instruments or lets critical ones drift too long.
- What is a reliability-based calibration interval?
- A reliability-based interval is set so a target in-tolerance rate is maintained, for example keeping at least 90 to 95 percent of calibrations in tolerance as-found. If too many come back out of tolerance, the interval is too long and you shorten it. If an instrument is reliably in tolerance calibration after calibration, the interval may be safely lengthened, saving cost and downtime. This is the most defensible method because the number comes from the instrument's own behavior. Recognized frameworks such as NCSLI RP-1 and ILAC-G24 describe the approach.
- Can you extend a calibration interval, and how do you defend it?
- Yes. Intervals are not set once. Lengthen them when the instrument's own history clearly supports it, and shorten them after an out-of-tolerance result, when the instrument is used harder, or when the measurement becomes more critical. The part that matters at audit is that each change has a recorded basis, so the interval is never an unexplained number. A calibration record carrying the interval, the basis, and the as-found and as-left readings answers what an assessor asks.
- Why watch drift instead of just pass or fail?
- An instrument that passes every calibration but drifts steadily toward its tolerance limit is telling you the interval is getting risky before it ever fails. Charting the as-found reading against the tolerance band over time surfaces that trend early, so you can shorten the interval before an out-of-tolerance event affects product. A pass or fail alone hides that movement entirely.
Put it into practice
Got a gage list? Send it over and we load it for you, usually in a couple of business days. Free on every plan. Then log every calibration to a tamper-evident audit trail and produce records on demand.
Axiospec is a documentation and workflow tool. It helps you keep clean, traceable, audit-ready records; certification depends on your own processes, scope, and assessor.
More like this, by email
A short series of practical calibration guides like this one. No spam, unsubscribe with one click.