We developed a multiple-regression model for estimating the weight contribution of prey taxa in fish diets. In this method, the counts of prey items are regressed against the total stomach content weight for individual fish in order to estimate the mean weight per individual prey item. Estimates obtained by this method were comparable to those obtained by direct measurement of the weights of individual prey items within each stomach. The multiple-regression method substantially reduced processing time by eliminating the need to sort and weigh individual prey items. Reductions in the labor required to process each stomach allow the investigator to examine a fixed number of fish at lower cost or to examine more fish during a fixed time.