Effects of nicotine, administered by continuous infusion via osmotic minipumps, were studied on the operant self-administration of alcohol by rats, using a variable interval (15 s) schedule, and measuring the acquisition, maintenance, extinction and reinstatement of responding for alcohol. Doses of nicotine of 0.25, 1.25 and 7.5 mg/kg/24 h had no significant effects on the maintenance of responding for alcohol, but 5 mg/kg/24 h nicotine resulted in a significant increase in responding on the lever delivering the reward when water was substituted for the alcohol, indicating delayed extinction of responding. During infusion of 2.5 mg/kg/24 h nicotine, responding was significantly greater over the "sucrose-fading" training sessions, during acquisition of responding, when mixtures of alcohol and sucrose were provided as reward. When minipumps infusing 2.5 mg/kg/24 h nicotine were implanted after the alcohol responding had been acquired, the responding for alcohol increase during the first week of nicotine infusion, but corresponding nicotine infusion doses of 0.25, 1.25 and 7.5 had no significant effects. The results indicate that nicotine can increase operant responding for alcohol and this is crucially dependent on the dose of nicotine and the time of testing. The results have implications for the frequently encountered dependence on the combination of alcohol and nicotine.