Rats were trained to respond with intravenous cocaine as the reinforcer under a fixed interval 15-min schedule, during which conditioned stimuli paired with cocaine were presented contingent on completion of a fixed ratio of 10 responses (i.e., second-order schedule of reinforcement). The effects of contingent and noncontingent cocaine were investigated. The results show that pretreatment with noncontingent (i.e., experimenter-administered) cocaine led to a satiation-like effect that was reflected in decreased numbers of responses and a tendency for an increased latency to initiate responding when the doses of cocaine administered were similar to or higher than the training/maintenance dose of cocaine. By contrast, noncontingent administration of cocaine doses lower than the training/maintenance dose, and response-contingent cocaine administration, led to increased drug-seeking behavior, as reflected in increased numbers of responses. The present data indicate that at least two factors determine whether administration of cocaine would lead to drug-seeking behavior: whether the cocaine administration is contingent or noncontingent, and the relative magnitude of the cocaine dose administered in relation to the training/maintenance dose of cocaine.