Rats were trained in a symbolic delayed matching-to-sample task to discriminate hedonic sample stimuli that consisted of food or no food. Retention functions decreased more rapidly on trials initiated by a food sample than on trials initiated by a no-food sample when retention intervals were manipulated within session. The asymmetrical functions could not be explained in terms of mediation of choice responding by magazine head-entry behavior during the retention interval. Unlike within-session changes in retention interval, between-session changes did not result in steeper forgetting functions for food samples. These results in rats are consistent with previous findings reported for the presence and absence of visual samples in pigeons (Wixted, 1993).