Rhesus monkeys, surgically prepared with intravenous catheters, were given opportunities to self-administer morphine for 3 days, methamphetamine for 2 days and saline for 2 days in a constantly repeating cycle. Access to drugs was limited to a 15-minute period every 4 hours. After stable base-line self-administration rates, saline or various concentrations of naltrexone were infused continuously through the catheter. In the first phase of the study each concentration of naltrexone was infused for 4 weeks (separated by 3 weeks of saline) while the dose of morphine available for self-administration was held constant at 8 microgram/kg/injection. Stable naltrexone dose-related suppression of morphine self-administration occurred throughout each 4-week infusion. In the second phase of the study, various doses of morphine were made available for self-administration during 6- to 8-week continuous infusions of saline or various concentrations of naltrexone. The dose-effect curve relating self-administration rate to morphine dose per injection shifted to the right and decreased in maximum as the rate of infusion of naltrexone increased. Methamphetamine and saline self-administration rates were unaffected by naltrexone.