Short answer
Accuracy concerns how close a measured value is to the true value of the quantity being measured. Precision concerns how closely repeated results agree with one another under stated conditions. A measuring system can produce tightly grouped readings that are consistently displaced from an appropriate reference. Those readings are precise, but the agreement among them does not establish accuracy. 1 2
For careful measurement work, a third term matters: trueness concerns how the long-run average compares with a reference value. The familiar “near the target versus tightly grouped” explanation often blends accuracy with trueness. This article uses the distinctions in the International Vocabulary of Metrology, VIM3. 1 3
On this page
At a glance
Scroll to compare all columns →
| Question | Accuracy | Precision |
|---|---|---|
| Main concern | Agreement of a measured value with a true value. | Agreement among repeated measured values. |
| What comparison is involved? | A value and the quantity’s true value; practical checks use references. | Results from repeated measurements under specified conditions. |
| Does a tight group prove it? | No; the group may be displaced. | A small spread is evidence of precision under those conditions. |
| Is one reading enough to show it? | A reading may be close, but establishing that needs suitable evidence. | One reading does not establish repeated agreement. |
| Numerical description | VIM3 treats accuracy as qualitative, not a quantity with its own numerical value. | Spread can be expressed using measures such as standard deviation. |
| Related concept | Trueness and measurement error. | Repeatability and reproducibility. |
What each thing is
Accuracy
Accuracy expresses closeness to the true value of the quantity being measured. A measurement with a smaller measurement error is described as more accurate. In practice, reference materials, standards, and other measurement information support the assessment; the reference itself should not be treated as free of uncertainty. 1 4
Precision
Precision describes the agreement among repeated measurements of the same or similar objects under specified conditions. Those conditions matter: repetitions by one operator on one instrument are different evidence from measurements made across instruments, operators, or laboratories. The VIM links precision to repeatability and reproducibility conditions. 2
Key differences
Consistency does not establish closeness to a reference. An instrument that repeatedly reads high can yield a small spread. Repeating that instrument’s reading demonstrates consistency, while comparison with an appropriate reference can reveal the offset. 1 2 3
A good average can hide a large spread. A set of readings distributed widely around a reference may have an average near it. That result does not make each individual reading close. The long-run average relates to trueness; the spread relates to precision. 2 3
Uncertainty describes more than repetition. Measurement uncertainty can include both components estimated from repeated results and components associated with standards, corrections, or other information. A small repeat-to-repeat spread therefore need not mean small overall uncertainty. 4
How to tell them apart
Ask what is being compared. Comparing repeated results with one another addresses precision. Comparing results with an appropriate reference helps assess error and possible systematic effects. To interpret either comparison, specify the measured quantity, procedure, conditions, and reference information. 1 2 3
A long decimal display is not a demonstration of either property. The practical questions are whether repetitions agree, whether the result agrees with a suitable reference, and how uncertainty was evaluated. Merely writing more digits answers none of those questions. This follows from the separate definitions of accuracy, precision, and uncertainty. 1 2 4
Where they overlap
Both help describe the quality of measurements, and both are relevant when evaluating a measuring process. A useful result needs more than repeatability alone: systematic effects and uncertainty also matter. The VIM explicitly relates accuracy to both precision and trueness while cautioning against using the terms as synonyms. 1 3 4
Edge cases
One reading can happen to be close. That does not establish how reliably the procedure will reproduce a close result. Precision requires information about repetitions under stated conditions. 2
Three readings are not an infinite series. A small sample can suggest an offset or spread, but VIM’s formal trueness definition refers to the average of infinitely many repetitions. Real assessments estimate behavior from finite evidence. 3
“Accuracy ±…” needs interpretation. Technical specifications sometimes use accuracy informally. Under VIM3, accuracy itself has no assigned numerical quantity value, so determine what the stated limit or numerical specification actually describes. 1
Why the distinction exists
Measurement practice needs to distinguish disagreement caused by variation between repetitions from disagreement with a reference. These patterns require different investigations. A shared vocabulary prevents a claim of consistency from silently becoming a claim of closeness to the intended value. 1 2 3
Common misconceptions
- “Precise means correct.” A stable offset can accompany closely grouped results. 2 3
- “A correct-looking average makes every reading accurate.” Individual results may be widely spread. 1 2
- “More decimal places prove accuracy.” They do not supply reference agreement. 1
- “Repeatability captures all uncertainty.” Other components can contribute. 4
Examples
Suppose a reference value is 100.0 units, with uncertainty small enough for this illustration. Readings of 102.0, 102.1, and 102.0 are closely grouped but offset from the reference. Readings of 98.0, 100.0, and 102.0 average 100.0 but show more spread. These invented numbers illustrate the distinction; three readings alone do not establish a complete uncertainty evaluation or a procedure’s long-run behavior. 1 2 3 4
Sources
Sources checked October 3, 2026. Scope: measurement terminology in VIM3, rather than the different use of precision in information retrieval or classification metrics.
- JCGM/BIPM, VIM3 — Measurement accuracy, 2.13. Definition, notes, and historical-usage annotation.
- JCGM/BIPM, VIM3 — Measurement precision, 2.15. Repetitions, conditions, and measures of spread.
- JCGM/BIPM, VIM3 — Measurement trueness, 2.14. Long-run average and systematic effects.
- JCGM/BIPM, VIM3 — Measurement uncertainty, 2.26. Components and reference-standard contributions.