Isoproterenol 7.5 and 15 microgram/kg prolonged the time spent in drinking of water-deprived rats whereas 30 microgram had no significant effect. No dose of isoproterenol affected punished responses in a conflict procedure which uses water as reinforcer. Punished responses were increased by 2.5, 5 and 10 mg/kg of diazepam but only the first two doses of diazepam increased drinking in normal deprived rats. The data show that the effects on punished responses and drinking can be dissociated and suggest that procedures using conditioned punishing of consummatory responses may be more appropriate than those using naive animals for revealing "true" antipunishment activity.