Feeding Snowy (Egretta thula) and Great (Casmerodius albus) egrets were observed during two breeding seasons in coastal New Jersey and two brief winter periods in northeast Florida. A number of tests based on assumptions of foraging models, predictions from foraging theory, and earlier empirical tests concerning time allocation and movement in foraging patches was made. Few of the expectations based on foraging theory and/or assumptions were supported by the empirical evidence. Snowy Egrets fed with greater intensity and efficiency during the breeding season (when young were being fed) than during winter. They also showed some tendency to leave patches when their capture rate declined, and they spent more time foraging in patches when other birds were present nearby. Great Egrets, in Contrast, showed few of these tendencies, although they did leave patches when their intercapture intervals increased. Satiation differences had some influence on feeding rates in Snowy Egrets, but only at the end of feeding bouts. Some individuals of both species revisited areas in patches that had recently been exploited, and success rates were usually higher after the second visit. Apparently, for predators of active prey, short—term changes in resource availability (_resource depression₎ may be more important than resource depletion, a common assumption in most OFT models.