What Is Calibration?

What calibration is

Calibration is the process of comparing a measuring instrument or reference material against a standard of known, documented accuracy, then recording how closely the instrument agrees with that standard. It is fundamentally a measurement of a measuring device: you take something whose “correctness” is in question — a balance, a thermometer, a pipette, a pH meter, a thermal cycler’s block temperature — and check its output against a reference whose value is already known to a tighter, traceable level of certainty.

The result of a calibration is not automatically a pass or fail. It is a data point: at this reading, on this date, under these conditions, the instrument’s output differed from the reference by this much. What happens next — whether that difference is small enough to leave the instrument in service, whether it needs adjustment, or whether it needs to be taken out of use — is a separate decision, made against a tolerance or acceptance criterion the lab sets in advance.

Why calibration exists as a formal practice

Every measuring instrument drifts. Mechanical wear, electronic component aging, temperature and humidity exposure, contamination, and ordinary use all shift an instrument’s output away from true value over time, usually gradually and invisibly to the person using it day to day. Calibration is the check that catches that drift before it silently corrupts data.

The problem calibration solves is trust: a measurement is only as good as the instrument that produced it, and a result is only defensible if there’s a documented basis for believing the instrument was reading correctly at the time. Without calibration records, “the balance read 4.998 g” is an unverifiable claim. With them, it’s traceable evidence.

Calibration applies wherever a measurement feeds a decision that matters — a diagnostic result, a formulation weight, a temperature-sensitive reaction, a regulatory filing, a safety threshold. That spans clinical and research laboratories, manufacturing and quality control, pharmacies, environmental monitoring, and calibration of the reference standards themselves at national metrology institutes.

NIST traceability and the calibration chain

A calibration is only meaningful if the reference used to perform it is itself trustworthy. That trustworthiness comes from metrological traceability: an unbroken chain of calibrations, each one linking a working reference back through progressively more accurate standards, ultimately to a national or international measurement standard — in the United States, standards maintained or recognized by the National Institute of Standards and Technology (NIST).

In practice, this is why calibration certificates and calibration weights are often described as “NIST-traceable”: the reference standard used in that calibration can be connected, step by step, back to NIST’s primary standards, with the measurement uncertainty at each link documented. A calibration performed against an uncalibrated or undocumented reference isn’t traceable, no matter how accurate the reference happens to be — traceability is about the documented chain, not just the underlying accuracy. For a detailed walkthrough of what a compliant certificate actually has to show and how “NIST-traceable” gets misused, see Calibration Certificates and Metrological Traceability.

Calibration vs. verification vs. qualification (IQ/OQ/PQ)

These three terms get used loosely in day-to-day lab conversation, but they describe distinct activities, and an auditor or accreditation assessor will expect them used correctly:

• Calibration compares an instrument’s readings to a traceable reference and quantifies the deviation. It answers “how far off is this instrument, and by how much.”
• Verification is a narrower check — often at a single point or a small number of points — confirming an instrument is still within an already-established tolerance, without necessarily generating the full traceable data set a calibration does. A verification is frequently used between full calibrations as an interim confidence check.
• Qualification (commonly structured as IQ/OQ/PQ — Installation Qualification, Operational Qualification, Performance Qualification) is a broader, documented exercise confirming that a piece of equipment was installed correctly, operates according to its specifications, and performs consistently under real conditions of use. Calibration is typically one component nested inside OQ or PQ, not a substitute for it: qualifying a thermal cycler, for example, includes calibrating its block temperature, but also verifying things calibration alone doesn’t touch, like lid function, ramp rate, and well-to-well uniformity.

For a fuller side-by-side comparison of when each term applies and how they interact in a validation lifecycle, see Equipment Qualification vs. Validation vs. Calibration.

Why calibration matters for data integrity and accreditation

Calibration is a foundational requirement, not a peripheral one, under the quality frameworks research and testing labs are commonly assessed against:

• ISO/IEC 17025 (the general standard for testing and calibration laboratory competence) requires accredited labs to establish metrological traceability for equipment that affects the validity of results, and to maintain calibration records an assessor can audit.
• Good Laboratory Practice (GLP, 21 CFR Part 58) requires that equipment used in nonclinical safety studies be adequately tested, calibrated, and standardized, with records retained as part of the study’s data integrity trail.
• Broader GxP and data-integrity expectations (often summarized as ALCOA+ — attributable, legible, contemporaneous, original, accurate, among other principles) treat an out-of-calibration instrument as a direct threat to the accuracy of every result it produced while out of tolerance, not just a maintenance lapse.

This is why calibration failures are a recurring finding in inspections and accreditation audits: a missed calibration interval, an undocumented adjustment, or a certificate missing required elements doesn’t just flag one instrument — it puts a question mark over every result that instrument generated since its last known-good check.

How often should equipment be calibrated?

There is no single universal interval. Calibration frequency is generally set based on manufacturer recommendation, the instrument’s observed drift/stability history, criticality of the measurements it supports, and how heavily it’s used — then documented and justified, not chosen arbitrarily. A lab typically starts with a manufacturer- or industry-typical interval, then lengthens or shortens it based on actual performance data (an instrument that consistently passes with a wide margin may move to a longer interval; one that drifts close to tolerance may move to a shorter one). Whatever interval is chosen, it needs to be written down and followed consistently — an assessor will ask why a given interval was chosen, not just whether it was met. For the mechanics of setting and justifying an interval, see Calibration Interval: How to Determine, Justify, and Document Re-Calibration Frequency.

Documenting a calibration

A calibration that isn’t documented is, for compliance purposes, indistinguishable from one that never happened. At minimum, a calibration record needs to identify the instrument, the reference standard used and its own traceability, the date, the environmental conditions if relevant, the as-found and as-left readings, the uncertainty of measurement, and a pass/fail determination against a stated tolerance. For labs seeking or maintaining ISO/IEC 17025 accreditation specifically, see ISO 17025 Calibration Certificates: Interpretation Guide for what a compliant certificate has to contain and how to read one from an outside vendor.

Who handles calibration in a research organization

Calibration responsibility is usually distributed rather than owned by one role:

• Lab managers and bench scientists typically own day-to-day scheduling, performing or arranging routine calibrations, and flagging equipment that fails a check.
• Quality assurance and compliance staff own the calibration program itself — the written procedure, interval-setting rationale, and audit-readiness of records — especially where ISO 17025 or GLP status is at stake.
• Procurement and equipment managers factor calibration into purchasing decisions: whether a new instrument ships with an initial calibration certificate, what an ongoing calibration contract will cost, and whether a vendor’s turnaround time fits the lab’s operating schedule.

Calibration and equipment procurement

Calibration requirements are also a purchasing consideration, not just an in-service one. Buying equipment without accounting for calibration means discovering the real cost and lead time only after the instrument is already on the bench. A procurement process for calibration-sensitive equipment should ask upfront: does the unit ship with a traceable initial calibration certificate, what’s the recommended interval, is an accredited calibration provider available locally, and what does a service contract cost over the equipment’s expected life. See Autoclave RFP Requirements and Buying a Used Thermal Cycler for worked examples of building calibration and validation requirements into an equipment purchase. For structuring a broader lab equipment RFP process, see What Is an RFP for Lab Procurement?, and for managing consumables and reference standards that support ongoing calibration work, see What Is Vendor-Managed Inventory for Lab Supplies?

Calibration and accreditation programs

Calibration performance also feeds directly into broader accreditation programs a lab or facility may be pursuing. In clinical laboratory settings, calibration and quality-control performance are core to CLIA compliance — see What Is CLIA? for an overview of that framework. In hospital and health-system settings more broadly, equipment maintenance and calibration records are part of what surveyors review under Joint Commission accreditation — see What Is Joint Commission Accreditation?

Related equipment-specific calibration guides

This page covers calibration as a general concept. For instrument-specific calibration requirements and procedures, see: Balance Qualification, Thermal Cycler Validation (IQ/OQ/PQ), Thermal Cycler Calibration, Analytical Balance Calibration & USP 41 Minimum Weight, and Out-of-Tolerance Calibration: What to Do When a Check Fails. If you’re deciding whether a piece of equipment is a scale or a balance in the first place — a distinction that affects what calibration standard applies — see the dictionary entry on scale vs. balance.

Frequently asked questions

What is the difference between calibration and verification?

Calibration is the full comparison against a traceable reference that quantifies deviation across a range of points and generates a documented record of the instrument’s actual performance. Verification is a narrower, often single-point check confirming an instrument is still within an already-established tolerance — it’s commonly used as an interim check between full calibrations, not a replacement for them.

What is the difference between calibration and qualification?

Qualification (IQ/OQ/PQ) is the broader exercise of confirming equipment is installed correctly, operates to specification, and performs consistently in real use. Calibration is typically one component performed as part of operational or performance qualification, not a substitute for the whole process.

What does “NIST-traceable” mean?

It means the reference standard used in a calibration can be connected through a documented, unbroken chain of calibrations back to a primary standard maintained or recognized by NIST, with the measurement uncertainty at each step recorded. It describes the documented chain, not simply that a reference is accurate.

How often should lab equipment be calibrated?

There’s no single universal interval. It’s typically set from manufacturer guidance and then adjusted based on the instrument’s actual drift history, how critical its measurements are, and usage level — and the chosen interval needs to be documented and justified, not just followed.

Who is responsible for calibration in a research organization?

It’s usually shared: lab managers and bench staff handle day-to-day scheduling and execution, quality assurance or compliance staff own the written calibration program and audit-readiness, and procurement/equipment managers factor calibration cost and logistics into purchasing decisions.