These experiments evaluated the ability of naltrexone (NTX) to reduce selectively oral and i.v. ethanol-reinforced responding, and examined the ethanol-NTX interaction in terms of the competitive opioid antagonist property of NTX. Five rhesus monkeys self-administered ethanol or sucrose and concurrently available water. Ethanol concentration was varied from 0.25% to 8% (w/v). Naltrexone (0.032-0.32 mg/kg) or saline was given i.m. 30 min prior to some drinking sessions. NTX (0.32 mg/kg) reduced ethanol-reinforced responding at the concentration that maintained the most responding (1% or 2%). NTX (0.1 mg/kg) reduced ethanol-reinforced responding, both at a low ethanol concentration (0.25%) that produced little ethanol intake (g/kg), and at a higher concentration (4%) with an appreciable intake. Thus, NTX (0.1 mg/kg) shifted the ethanol concentration-consumption curve down, in an insurmountable manner. NTX (0.1 and 0.32 mg/kg) also reduced reinforced responding for sucrose 100 g/l. In another experiment, three rhesus monkeys were given opportunities to self-administer ethanol i.v. NTX (0.1 mg/kg) reduced the number of ethanol injections obtained by the monkeys at all ethanol doses tested (0.01, 0.032, and 0.1 g/kg per injection). The dose-effect curve was also shifted down. These results showed that NTX reduced behavior maintained by either ethanol or sucrose non-selectively. Furthermore, the ability of NTX to suppress ethanol-reinforced responding did not depend on the route of ethanol administration and was not overcome by increasing the concentration or dose per injection of ethanol.