Lever pressing by rats was maintained under a fixed-ratio 30 schedule of food presentation. The response rate-decreasing effects of several opioid compounds that vary in selectivity for, and activity at, µ, kappa, and non-opioid receptors, were examined alone and in combination with the opioid antagonist naltrexone. Naltrexone (0.01-1.0mg/kg) produced dose-dependent and generally parallel rightward shifts in the dose-effect curves for morphine, fentanyl, butorphanol and nalbuphine. Apparent pA(2) values for naltrexone against these agonists ranged from 7.05 to 7.29, and the slopes of the regression lines fitted to the Schild plots approximated theoretical unity (-1.0), suggesting a competitive interaction at µ-opioid receptors. In contrast, although at least one dose of naltrexone (0.01-10.0mg/kg) antagonized the response rate-decreasing effects of bremazocine, U50,488, (-)-pentazocine and nalorphine, suggesting some opioid activity, these effects differed from those of the µ agonists in that: (a) they were less sensitive to naltrexone antagonism; (b) maximal rightward shifts were smaller; (c) antagonism patterns were not directly related to naltrexone dose and, in some cases, were influenced by the response rate-decreasing effects of the larger naltrexone doses; and (d) there were considerable between-subjects differences in sensitivity to naltrexone antagonism. Naltrexone (0.1-10.0mg/kg) did not antagonize the effects of the non-opioid control compound pentobarbital. The present results suggest that patterns of naltrexone antagonism can provide a basis for making inferences about receptor activity related to the effects of some opioids on rate of schedule-controlled behavior. With some opioids, however, such inferences are limited by the direct response rate-decreasing effects of naltrexone itself and by differences in patterns of antagonism across subjects.