Rats were trained to discriminate 3.0mg/kg morphine from water in a standard two-lever drug discrimination procedure and tested in a hot water tail-withdrawal procedure. When tested with morphine, fentanyl, buprenorphine and butorphanol, individual animals showed a three- to ten-fold difference in the lowest dose that substituted completely for morphine, whereas 30- to 1000-fold differences were obtained with nalbuphine and levallorphan, respectively. Across repeated determinations, the dose-effect curves for morphine and nalbuphine remained relatively stable within an individual, suggesting that the profound individual differences with nalbuphine were not a consequence of variability in the dose-effect curve. In addition, despite the extremely shallow group dose-effect curves obtained with nalbuphine and levallorphan, intermediate levels of drug-appropriate responding were not evidenced in individual animals. In the tail withdrawal procedure, the doses of morphine and fentanyl required to produce the maximal levels of antinociception varied by approximately three-fold across individual rats. With butorphanol and nalbuphine, differences across animals were greater than 30-fold and 300-fold, respectively, whereas with levallorphan, substantial individual differences were observed in the maximal level of antinociception. Further analyses indicated that animals sensitive to the stimulus effects of nalbuphine and levallorphan were also sensitive to the antinociceptive effects of nalbuphine, levallorphan and butorphanol. In contrast to the individual differences obtained with the stimulus and antinociceptive effects of these opioids, the potency of these drugs for decreasing rate of responding was similar across animals. These findings indicate that the relative efficacy of an opioid at the mu receptor is an important determinant of individual differences in responsiveness to its stimulus and antinociceptive effects, but not to its rate-decreasing effects.