By using either water or food reinforcement, rats were trained to perform a discriminated lever release task (DLR), which required the rat to hold an operant lever in the closed position through one of five randomly presented foreperiods (2-6 s) and to release the lever within 0.5 s of the onset of a light discriminative stimulus. The procedure is analogous to the method used in human reaction time studies, except that here the procedure was free-operant, not fixed-trial. The effects of pimozide (0.12, 0.25, and 0.50 mg/kg) on this behavior were evaluated in terms of numbers of total responses, reinforced responses (successful releases), anticipatory responses, and extended responses (holding too long). Significant dose-dependent decreases in total responses and in reinforced responses were seen as supporting the hypothesis of a deficit in response initiation, which is often invoked to account for neuroleptic-induced reductions in discriminated active avoidance. Pimozide also increased the proportion of extended responses, suggesting that the drug affected the nature of responding as well as the tendency to respond. In the DLR task, pimozide produced substantial within-session decrements in both total responses and number of reinforced responses; however, extended responses exhibited within-session increases at the lowest dose. The results were discussed from both behavioral and pharmacological perspectives. The former emphasized motor effects and response initiation deficits, while the latter jointly considered neuronal responses to neuroleptic challenge and the dopamine release that results from behavioral activity itself.