Rats were trained and extinguished under conditions that were comparable in all respects except for the reinforcement proc edures. Food 5s responded more frequently during ex tinction than IC5 animals, whether the training had been with CRF or FR-lO reinforcement schedules. In an attempt to ascertain if behavior reinforced with intracranial stimulation (ICS) differs from more tradi tionally reinforced behavior, such as food-reinforced behavior, a variety of comparison studies have been undertaken. One such area of research has been con cerned with the comparison of resistance to extinction of a response following ICS or food reinforcement. Initially, Olds & Milner (1954, p. 425) reported that stimulation of the brain ... produces acquisition and extinction curves which compare favorably with those produced by a conventional reward. However, further observations of the persistence of ICS-reinforced re sponding led to the general impression (cf., Kimble, 1961: Morgan, 1965) that such Ss were less resistant to extinction than food-reinforced animals. One inter pretation of these differences in persistence emphasized the variations in the procedures utilized in delivering reinforcement in ICS and food experiments. For example, several studies have examined the effects of modifying the reinforcement contingencies in ICS situations (e.g., Gibson et aI, 1965; Pliskoff et aI, 1965). In general, these studies conclude that response persistence following ICS reinforcement is not necessarily inferior to that produced by food reinforcement. The present study was undertaken to determine if differences in free operant responding and subsequent resistance to extinction do in fact exist when food and ICS reinforcement contingencies are not manipulated, but rather, when several other sources of uncontrolled variation are held constant. Some variables which could distort comparisons of food and ICS reinforcement in the typical experiment are the long-term effects of pento barbital sodium anesthetic on the acquisition and reten tion of ICS animals (e.g., Stern, 1960): the restraining effects exerted by the electrode lead (Seward et aI, 1959): and the differential effectiveness of food-related cues for deprived and non-deprived animals (Campbell & Sheffield, 1953). The present study attempted to control the above variables in order to provide a direct comparison of the effects of food and ICS reinforcement procedures. Method