Even the most experienced geotechnical engineer is likely to assume that the results of cone penetration tests are unquestionably accurate, reliable and repeatable. There are, however, multiple factors, some that have nothing to do with the soil properties, that need to be carefully addressed prior to testing if the equipment is to return results that can be relied on for design purposes. In soils which are very soft or soft, cone penetration test results can be particularly sensitive to the method of calibration. A high degree of rigour to the calibration process is required, otherwise there is a risk that the results obtained could be inaccurate and adversely on the reliability of the interpretation of design soil strength profiles. In this technical note sources of error in cone calibration are discussed. Reference is made to ISO 22476-1 which was revised in 2022, with the addition of a defined approach to calibration. Examples are used to demonstrate the typical errors that could be introduced during calibration.