Lever pressing by squirrel monkeys was maintained by i.v. injections of nicotine (3-560 microgram/kg) or cocaine (3-300 microgram/kg) under two intermittent schedules of self-administration. Under a fixed-interval schedule, the first response after a specified interval of time produced an injection. Under a second-order fixed-interval schedule, the completion of every 10-response fixed-ratio unit produced a brief visual stimulus and the first fixed-ratio unit completed after a specified interval produced both the brief stimulus and an injection. As the dose of either drug was increased, the rate of responding first increased and then decreased; maximal response rates maintained by nicotine were approximately equal to those maintained by cocaine in some monkeys, but less than those maintained by cocaine in other monkeys. Patterns of responding maintained by the two drugs were qualitatively similar in all monkeys and were characteristic of performances maintained by other reinforcers under fixed-interval or second-order fixed-interval schedules. Doses of nicotine greater than 30 microgram/kg/injection usually produced vomiting, but often maintained responding well above the rates maintained by saline. When the nicotinic antagonist, mecamylamine (1.0 mg/kg i.m.) was administered before every experimental session, responding maintained by nicotine, but not by cocaine, fell to within saline-control levels; increasing the dose of nicotine to as high as 1700 microgram/kg/injection did not restore responding. Under the intermittent schedules studied here, nicotine served as an effective reinforcer to maintain responding and the reinforcing effects of nicotine were blocked by mecamylamine.