Pigeons were trained to peck a response key that briefly produced stimuli correlated with the passage of time in a fixed-interval schedule of food presentation for pecks on another response key. Pecks on the key that produced food were not likely near the end of each fixed interval, whereas pecks on the key that produced the discriminative stimuli were most likely to occur during the middle portion of each fixed interval. Chloropromazine (3.0--30.0 mg/kg) and d-amphetamine (0.03--3.0 mg/kg) produced inverted U-shaped dose-effect curves with response rate on the food-producing key as the dependent variable, and monotonically decreasing functions were obtained for the discriminative-stimulus producing responses. A "rate-dependency" interpretation adequately described the drug effects on the temporal distributions of pecks on the food-producing key, but was not consistent with the disruption of the temporal distribution of pecks on the key that produced the discriminative stimuli. The pecks on the discriminative-stimulus producing key had some of the properties of schedule-induced, or adjunctive, behavior. The effects of amphetamine and chlorpromazine on the rate of adjunctive behaviors may not be predictable from control rates.