Amphetamine was administrated intramuscularly 20-25 min before each experimental session in doses of: 0.2 mg/kg, 0.5 mg/kg and 1.0 mg/kg. Each dose was applied twice and preceded by a regular experimental sessions without any treatment. An instrumental performance consisted of alimentary-social differentation of two tones reinforced either by food or by sensory-social reward (petting by the experimenter). Amphetamine produced dose dependent decrease of the instrumental performance of both alimentary and social responses. This decrease was however not equal regarding both reactions. The dose of 1 mg/kg produced deep, statistically significant deterioration of alimentary as well as social responses. After the administration of the dose of 0.5 mg/kg the decrease of alimentary responses was equal to that produced by 1 mg/kg, whereas social responses was equal to that produced by 1 mg/kg, whereas social responses were less deteriorated. After the dose of 0.2 mg/kg the reduction of alimentary responses was smaller than produced by 0.5 mg/kg but still more pronounced than in the case of socially reinforced reactions. The results confirm our previous data that amphetamine suppresses positively motivated instrumental performance in dogs. The results also show that amphetamine-evoked suppression is dose dependent and that it is different for alimentary and social responses. This might indicate that the positive reward system is not homogenous but consists of some subsystems related to different kinds of reward. It is concluded that the suppressing effect of amphetamine is due to the inhibitory effect on motivation.