The studies presented here were designed to further clarify the nature of nicotine self-administration (SA) based on a limited access model in which rats are food restricted, receive operant training using food reinforcement, and are then tested in daily 1-h drug sessions. We examined the effects of dose, feeding schedule, and contingency of drug delivery on acquisition of nicotine SA. Two doses of nicotine bitartrate, 0.03 and 0.06 mg/kg per infusion (free base), supported the transition from food-reinforced to drug-reinforced responding, although the pattern of behavior differed between these doses. In contrast, 0.01 mg/kg per infusion failed to maintain nicotine SA. In a second study, animals were divided into three groups according to feeding schedule. Rats that were both weight restricted and food deprived showed the highest level of SA behavior, although neither food deprivation nor weight restriction was necessary to establish SA. In the third experiment, rats that were switched from food to nicotine as the response-dependent reinforcer maintained higher response rates throughout a 9-day period than animals switched to response-independent (i.e., yoked) nicotine which showed minimal responding after day 1. Furthermore, the differences between self-administering and yoked animals emerged during the first session, suggesting that nicotine may serve as a reinforcer during the first drug exposure in naive animals. These results indicate that acquisition of nicotine SA can be influenced by both dose of nicotine and feeding schedule and that, in animals previously trained on a food-reinforced operant, active lever pressing is maintained only when nicotine delivery is contingent upon responding.