Previous studies of toxicant effects on fish foraging behavior and predator‐prey interactions have taken a strictly empirical approach. The most common observation of altered feeding behavior was cessation of feeding or reduction in the amount of artificial food consumed. Changes in the number of live prey attacked and captured, latency to feed and capture efficiency have also been documented. Predator‐prey tests have placed major emphasis on toxicant effects on the ability of prey to escape predation. Several different test systems, as well as model ecosystems, have been used. It is difficult to evaluate the sensitivity of these behavioral tests in relation to standard chronic tests. There was no consistency in test design, and few studies provided information on chemical effects on reproduction or growth in conjunction with behavioral effects. These empirical studies have little basis in ecological theory and do not allow development of testable hypotheses a priori to field verification of laboratory results. Few laboratory feeding studies have been verified in the wild. A mechanistic approach to feeding studies using optimal foraging and bioenergetics models may provide sensitive tests of contaminant effects that may be readily verified in the field. Model simulations demonstrate how toxicant effects on components of fishes' predation sequence can modify the size‐frequency distribution of prey in the fishes' diet and how reductions in the amount of food consumed may alter growth.