The use of irreversible antagonists to assess opioid efficacy has proven fruitful for classifying opioids on the basis of high or low efficacy, but few studies have provided quantitative estimates of efficacy. The purpose of this study was to use -funaltrexamine (-FNA) and clocinnamox (C-CAM) in a drug discrimination procedure to examine the efficacy of fentanyl, morphine, l-methadone, sufentanil, and etorphine. In pigeons trained to discriminate 0.12 mg/kg fentanyl from water, dose-effect curves were determined for each opioid alone and after pretreatment with -FNA and C-CAM. Using quantitative analyses according to an extended model of Black and Leff (1983), apparent efficacy () and affinity (K A ) of each opioid was determined, as well as the degree of receptor inactivation (q) produced by each dose of each antagonist. -FNA and C-CAM produced dose-and time-dependent, rightward shifts in the dose-effect curves of each opioid, and analyses based on dose-ratios and values suggest a rank order of efficacy of etorphine Ͼ sufentanil ϭ l-methadone Ͼ fentanyl ϭ morphine. Marked differences in the profiles of antagonism produced by -FNA and C-CAM were also apparent, as C-CAM, but not -FNA, produced insurmountable antagonism. The q values for each antagonist were consistent with these data in indicating that C-CAM and -FNA can inactivate nearly 100 and 75% of the receptor population, respectively. In tests conducted in pigeons chronically treated with morphine, doses of -FNA that produced parallel, rightward shifts in untreated pigeons flattened the morphine dose-effect curve in morphine-treated pigeons. These results indicate that -FNA and C-CAM can differentiate opioids with high relative efficacy and yield comparable estimates of efficacy for various opioids. There are, however, limitations in the proportion of the receptor population that can by eliminated by -FNA.