A pipette display states a setting. Calibration provides evidence about the relationship between that setting and delivered volume under specified conditions. The two should not be treated as interchangeable entries in a sample record.
Ask what the calibration actually assessed
Eppendorf’s calibration guide distinguishes calibration, which establishes the relationship between selected and dispensed volume, from adjustment, which changes the dispensing system. It also treats the device and accessories, including the tip, as a system.Eppendorf — Calibration and adjustment of dispensing systems in the laboratory (opens in a new tab)
A label saying “calibrated” therefore leaves useful questions unanswered. Which device was tested, at which settings, using which configuration, and what results were obtained?
The record should allow the pipette used in an experiment to be connected to its own calibration information. A certificate for another device of the same model is not evidence about the specific instrument used.
An adjustment date and a calibration result have different meanings. If the device was adjusted, retain the relevant before-and-after information rather than assuming that the adjustment itself demonstrates the final performance.
Separate average offset from variation
The manufacturer guide distinguishes systematic deviation from random variation in a series of delivered volumes. A tightly grouped series can still be centred away from the selected value.Eppendorf — Calibration and adjustment of dispensing systems in the laboratory (opens in a new tab)
| Observed pattern | Question it raises |
|---|---|
| Readings tightly grouped near 98 µL | Is there a consistent offset from the setting? |
| Readings widely scattered around 100 µL | How variable are individual deliveries? |
| Mean near 100 µL with limited scatter | Do the assessed performance and conditions meet the intended requirement? |
These patterns are explanatory examples, not pass/fail criteria. The study’s requirements and an appropriate measurement assessment determine whether the performance is suitable.
Do not report only a favourable mean if the spread matters to the experiment. Conversely, small scatter alone cannot establish that the average delivered quantity matches the setting.
Consider the gap between testing and use
Eppendorf’s guide notes that liquid properties, including density and vapour pressure, can affect air-cushion dispensing relative to the water used for adjustment. A water-based calibration is not automatic evidence for every liquid.Eppendorf — Calibration and adjustment of dispensing systems in the laboratory (opens in a new tab)
The practical documentary question is whether the experimental transfer falls within the assessed use. Record the selected setting and the relevant sample context without relabelling that setting as a separately measured delivery.
For example, “100 µL setting, device P4” accurately describes a recorded action. “Exactly 100.000 µL delivered” adds precision and certainty that the setting alone does not support.
Retain enough information to assess the volume
Keep the device identity, relevant calibration record, selected volume and any documented departure from the method connected to the experiment. These links help distinguish instrument evidence from assumptions about the transfer.
If a subsequent check identifies a problem, assess which earlier work may be affected using the actual use history. It does not follow that every historical transfer had the same error.
The aim is a volume statement that can be evaluated: what was selected, what was tested and how the experiment relates to that evidence. This is interpretation of records, not a calibration or pipetting protocol.
Sources and further detail
- Eppendorf — Calibration and adjustment of dispensing systems in the laboratory (opens in a new tab)
Manufacturer user guide checked for calibration versus adjustment, systematic/random deviation and liquid-property effects. Historical numerical standard limits and testing frequencies are not presented as current requirements.
Sources checked 20 September 2026. Numerical examples are hypothetical unless attributed to a study. Measurement explanations are not laboratory protocols or product handling limits. This article has not undergone independent scientific peer review.