Selected results from strain measurements on four nuclear pressure vessels are presented and discussed. The measurements were made in several different regions of the vessels: transition zones in vessel heads, flanges and bottom parts, nozzles, internal vessel structure and flange bolts. The results presented are based on data obtained by approximately 700 strain-gauges, and a comprehensive knowledge of the quality obtained by such measurements is established. It is shown that a thorough control procedure before and after the test as well as a detailed knowledge of the behaviour of the signal from the individual gauges during the test is necessary. If this is omitted, it can be extremely difficult to distinguish between the real structural behaviour and a malfunctioning of a specific gauge installation. In general, most of the measuring results exhibit a very linear behaviour with a negligible zeroshift. However, deviations from linear behaviour are observed in several cases. This nonlinearity can be explained by friction (flange connections) or by gaps (concentrical nozzles) in certain regions, whereas local plastic deformations during the first pressure loadings of the vessel seem to be the reason in other regions.