Recent studies have reported differences in the receptor mechanisms and intrinsic efficacies of heroin and its metabolites 6-acetylmorphine and morphine in rodents. The present study examined the generality of these findings to rhesus monkeys using two behavioral procedures. In an assay of schedule-controlled behavior, response rates were recorded under a fixed-ratio 30 schedule of food presentation. In an assay of thermal nociception, tail-withdrawal latencies were measured from warm water (42-58 degrees C). Heroin, 6-acetylmorphine and morphine produced dose-dependent rate-decreasing and antinociceptive effects. Antagonism studies were conducted with the competitive mu-selective antagonist quadazocine, the competitive delta-selective antagonist naltrindole, and the irreversible mu-selective antagonist beta-funaltrexamine (beta-FNA). Quadazocine dose-dependently antagonized the effects of all three opioids. Quadazocine pA2 values were similar across drugs and assays (7.4-7.8) and similar to quadazocine pA2 values for antagonism of other mu agonists. In contrast, naltrindole did not alter the effects of any of the opioids. beta-FNA antagonized the rate-decreasing and antinociceptive effects of heroin and morphine. Dose-effect data for heroin- and morphine-induced antinociception alone and after beta-FNA treatment were used to estimate in vivo apparent efficacy values (tau). Tau values (95% confidence limits) were 8.1 (6.9-9.6) for heroin and 2.6 (2.5-2.9) for morphine, but this difference is relatively small. These results suggest that the rate-decreasing and antinociceptive effects of heroin, 6-acetylmorphine and morphine are mediated by pharmacologically similar populations of mu opioid receptors in rhesus monkeys. The in vivo apparent efficacy of heroin at mu receptors was similar to or only slightly greater than that of morphine.