Substitution and antagonism patterns of butorphanol, meperidine and profadol were examined in pigeons trained to discriminate either a 0.056 (low) or 0.18 (high) mg/kg dose of fentanyl from saline. In the low dose group, fentanyl, meperidine, profadol and butorphanol substituted completely (≥ 85% fentanyl-appropriate responding) for the fentanyl stimulus. For fentanyl and butorphanol, complete substitution was obtained at doses that had little effect on rates of responding, whereas profadol substituted at doses that moderately decreased rates and meperidine substituted at doses that markedly decreased rates. Although naloxone antagonized the stimulus effects of meperidine and profadol, it failed to alter their rate-decreasing effects. In the high-dose group, fentanyl and butorphanol substituted completely and meperidine and profadol partially (approximately 50% fentanyl-appropriate responding) for the fentanyl stimulus. During antagonism tests, meperidine and profadol produced only small decreases (less than 15%) in the percentage of fentanyl-appropriate responding produced by the training dose of fentanyl. Analysis of individual data indicated that meperidine and profadol produced three patterns of substitution and antagonism. In one subgroup of pigeons, meperidine and profadol substituted completely for but failed to antagonize the fentanyl stimulus; in another, these opioids failed to substitute for but did antagonize the fentanyl stimulus; and in another, these opioids failed to substitute for or antagonize the fentanyl stimulus. In this latter subgroup, meperidine and profadol produced leftward shifts in the dose effect function for the stimulus effects of fentanyl and butorphanol. The present findings suggest that the failure of meperidine and profadol to substitute completely for the high-dose fentanyl stimulus was a direct consequence of their rate-decreasing effects rather than low efficacy at the mu receptor.