Cocaine-associated cues can elicit incentive motivational effects that drive cocaine-seeking behavior and contribute to relapse. The extinction/reinstatement model is commonly used to measure these effects in animals. This study examined the influence of training and testing schedules of reinforcement on cue-elicited reinstatement. Lever presses during training resulted in cues and cocaine (0.75 mg/kg/IV) on either continuous or partial reinforcement schedules [fixed ratio (FR) 1 or 11, variable ratio (VR) 5 or 11]. Animals then underwent extinction training, followed by a test for cue-elicited reinstatement of extinguished cocaine-seeking behavior by response-contingent cue presentations on either a continuous (FR 1) or a partial reinforcement schedule (FR 11). Partial reinforcement during training resulted in higher response rates during cue-elicited reinstatement relative to continuous reinforcement. In contrast, delivery of cues on a continuous reinforcement schedule during testing yielded higher response rates relative to delivery on a partial reinforcement schedule. Finally, the shift from a partial to a continuous reinforcement schedule across training and testing phases did not alter response rates. These findings provide important information for choosing parameters for reinstatement of drug-seeking behavior that would allow the most sensitive method to detect changes in response rate after an experimental manipulation.