Responding by squirrel monkeys was maintained under a schedule of i.v. nicotine postponement. Each lever-pressing response postponed an impending i.v. injection for 60 sec; in the absence of responding, injections occurred very 20 sec. The lowest dose of nicotine (10 micrograms/kg/injection) maintained rates of responding that were only slightly higher than those maintained by saline. Intermediate doses (30 or 56 micrograms/kg/injection), however, maintained rates and patterns of responding that were similar to those maintained by electric shock under an identical schedule of postponement. At the highest dose (100 micrograms/kg/injection), nicotine disrupted responding and induced vomiting in each monkey. Presession treatment with mecamylamine (1.0 mg/kg i.m.), a nicotinic antagonist with both central and peripheral effects, shifted the nicotine dose-response curve to the right by about 1 log unit and blocked nicotine-induced emesis. In contrast, presession treatment with hexamethonium (1.0 or 10.0 mg/kg i.m.), a nicotinic antagonist with primarily peripheral effects, did not produce an overall rightward shift in the nicotine dose-response curve, although it did reduce the rate of responding maintained by the intermediate doses of nicotine and blocked emesis. Neither mecamylamine nor hexamethonium systematically altered responding maintained by electric shock under the schedule of postponement. The results show that schedule-characteristic performances can be maintained by postponement of i.v. injections of nicotine and that the maintenance of responding under the postponement schedule depends largely on the actions of nicotine in the central nervous system.