The aim of the present experiment was to study the potentially discriminable effects of combinations of morphine and naltrexone during long-term treatment. Three groups of gerbils had to discriminate the effects of morphine (12 mg/kg) and those of either saline (4 ml/kg), naltrexone (2 mg/kg), or a combination of this dose of morphine plus naltrexone injected IP 60 min prior to the start of the discriminative training in a T-shaped maze. Rapid development of drug discriminative control of choice behavior (left or right turn in the maze) was evident in these 3 groups which is in marked contrast to the performance of gerbils trained with morphine-naltrexone combination vs. saline or gerbils trained with naltrexone only vs. saline. Neither of these latter groups reached the criterion of performing 8 correct first-trial choices in 10 consecutive training sessions during the 60 training sessions allowed, while the 3 other groups began their criterion performance after only 7--8 training sessions. Thus the discriminative properties of certain combinations of morphine and naltrexone are weak and therfore are not easily discriminable from the effects induced by saline.