The effects of nicotine, heroin, mecamylamine, and naltrexone were studied in rats trained to respond under a fixed-interval 3-min schedule of food presentation. Nicotine (0.1-3.0 mg/kg) first increased, then decreased response rates; heroin (0.03-0.6 mg/kg) produced only dose-related response rate decreases. Mecamylamine (0.1-3.0 mg/kg) and naltrexone (0.3-10.0 mg/kg), administered alone, each had little effect on response rates. However, when administered in combination with increasing doses of nicotine, mecamylamine (1.0 mg/kg) blocked the increases in response rate caused by 0.3 and 1.0 mg/kg nicotine and partially reversed the decreases in rate caused by 3.0 mg/kg nicotine. In contrast, the combination of naltrexone and nicotine, at doses of each that alone either had no effect or increased response rates, markedly decreased responding. This phenomenon was not evident during the first pairing of naltrexone and nicotine, but appeared in the second and subsequent pairings. After drug combinations had been tested, the nicotine dose-response curve was unchanged from its previous values, and naltrexone alone produced no tendency to decrease response rate. These observations may be related to previous results that have suggested a role for endogenous opioids in mediating certain of the behavioral effects of nicotine.