The accuracy of five different methods used in spraint analysis for extimating prey proportions is tested through a feeding experiment with three otters in captivity. Estimated prey proportions are copared to the actual diet of the otters. The results of analysing 362 spraints and the time budget indicated a score-bulk estimate to be the most appropriate method of estimating prey diet and it is concluded that dietary studies using different methods can be compared in broad outlines. The relatins between the methods are confirmed in the analysis of 978 spriaints, collected in the field.