Abstract: Saddle-fit is recognised as an important factor in the pathogenesis of back problems in horses and is empirically being evaluated by pressure measurements in clinical practice, although not much is known about the validity, reliability and usability of these devices in the equine field. This study was conducted to assess critically a pressure measurement system marketed for evaluating saddle fit. Validity was tested by calculating the correlation coefficient between total measured pressure and the weight of 28 different riders. Reliability and discriminative power with respect to different saddle fitting methods were evaluated in a highly standardised, paired measurement set-up in which saddle-fit was quantified by air-pressure values inside the panels of the saddle. Total pressures under the saddle correlated well with riders’ weight. A large increase in over-day sensor variation was found. Within trial intra-class correlation coefficients (ICCs) were excellent, but the between trial ICCs varied from poor to excellent and the variation in total pressure was high. In saddles in which the fit was adjusted to individual asymmetries of the horse, the pressure measurement device was able to detect correctly air-pressure differences between the two panels in the back area of the saddle, but not in the front area. The device yielded valid results, but was only reliable in highly standardised conditions. The results question the indiscriminate use of current saddle pressure measurement devices for the quantitative assessment of saddle-fit under practical conditions and suggest that further technical improvement may be necessary.