The purpose of these experiments was to examine the relationship of agonist relative efficacy to the pattern of tolerance and cross-tolerance to the morphine-like stimulus effects of three opioid agonists. Rats were trained to discriminate 3.2 mg/kg morphine from saline under fixed-ratio 15 schedule of food reinforcement. Morphine, nalbuphine, and fentanyl produced dose-dependent increases in morphine-like stimulus effects and decreases in response rates. Repeated treatment with 20 mg/kg per day morphine increased the ED50 for stimulus control by fentanyl, morphine, or nalbuphine two-, four-, or 40-fold, respectively. Repeated treatment with 64 mg/kg per day nalbuphine increased the ED50 for stimulus control for morphine by two-fold, but lower or higher treatment doses had no significant effect. Treatment with 100 mg/kg per day nalbuphine increased the ED50 for nalbuphine by six-fold. Repeated treatment with 0.22 mg/kg per day fentanyl increased the ED50 for stimulus control by fentanyl or morphine by approximately two-fold. Comparisons among treatment conditions suggested that magnitude of tolerance to morphine-like stimulus effects did not vary as an inverse function of the relative efficacy of the agonist used for repeated treatment. Rather repeated morphine and fentanyl treatments produced comparable tolerance, whereas repeated nalbuphine treatment did not evoke substantial tolerance. Comparisons within treatment conditions, however, suggested that magnitude of tolerance may vary inversely with relative efficacy of the agonist tested for morphine-like stimulus effects. During treatment with morphine or fentanyl, greater tolerance was observed to the morphine-like stimulus effects of the lower efficacy agonist relative to the higher efficacy agonist.