AbstractAs reported by Gauvin and Young [1987], a robust discrimination can be established among saline, 10 mg/kg morphine, and 1.8 mg/kg morphine. The present experiment further examined the pharmacological characteristics of this discrimination. Pigeons were trained to discriminate among i.m. injections of saline, 1.8 mg/kg morphine (the low‐dose, or LD, stimulus), and 10 mg/kg morphine (the high‐dose, or HD, stimulus). Performance was maintained under FR 30 schedules of food delivery in 30‐min experimental sessions. After establishment of stimulus control, various doses of morphine, the agonist‐antagonist nalbuphine, and the nonopioid pentobarbital were tested for generalization to the LD and HD stimuli. For morphine itself, generalization of the LD and HD stimuli varied as a function of dose. A dose of 0.10 mg/kg morphien occasioned responses to only the saline key. As the dose of morphine was increased to 1.8 mg/kg, the proportion of responses to the LD key increased from 0 to 100%. As the dose was increased further, the proportion of responses to the LD key decreased, with a commensurate increase in responses to the HD key. Complete HD generalization occurred at 5.6 or 10 mg/kg morphine. In contrast, the opioid agonist‐antagonist nalbuphine (1.0 to 56 mg/kg) evoked generalization to only the LD stimulus. The highest dose of nalbuphine markedly suppressed response rates. The nonopioid pentobarbital (0.1 to 18 mg/kg) evoked neither LD nor HD generalization. This pattern of generalization suggests that the discrimination among saline and two doses of morphine may provide a fruitful assay for the agonist efficacy of nove