AbstractXylem resistance to water stress‐induced cavitation is an important trait that is associated with drought tolerance of plants. The level of xylem cavitation experienced by a plant is often assessed as the percentage loss in conductivity (PLC) at different water potentials. Such measurements are constructed with samples that are excised underwater at native tensions. However, a recent study concluded that cutting conduits under significant tension induced cavitation, even when samples were held underwater during cutting. This resulted in artificially increased PLC because of what we have termed a ‘tension‐cutting artefact’. We tested the hypothesized tension‐cutting artefact on five species by measuring PLC at native tension compared with after xylem tensions had been relaxed. Our results did not support the tension‐cutting artefact hypothesis, as no differences were observed between native and relaxed samples in four of five species. In a fifth species (Laurus nobilis), differences between native and relaxed samples appear to be due to vessel refilling rather than a tension‐cutting effect. We avoided the tension‐cutting artefact by cutting samples to slightly longer than their measurement length and subsequent trimming of at least 0.5 cm of sample ends prior to measurement.