These experiments sought to determine whether dopamine (DA) could reverse the depressive effects of alpha-methyl-p-tyrosine (AMPT) on a conditioned avoidance response (CAR). Rats were randomly allocated to shocked groups (CAR-trained) and non-shocked (CAR-naive) groups. The CAR-trained rats, conditioned to avoid an electric shock, were administered AMPT (150 mg/kg at -24 h and 50 mg/kg at -1 h, both IP), nialamide (80 mg/kg IP at -1 h) and saline (1 microliter) or DA (5 or 10 micrograms/microliters, dissolved in 1 microliter saline, at time 0) directly into the nucleus accumbens. The rats were then tested for CAR at 0.5, 1, 2, 3, 4, 8, 12, 24 an 48 h. The CAR-naive rats, conditioned to the behavioural environment without electric shock being presented, were administered AMPT, nialamide and DA or saline as above. Both doses of DA antagonised the AMPT-induced suppression of the CAR in the CAR-trained rats, reaching a maximum 2-4 h after its local application. In the CAR-naive rats, DA produced a "pseudo-CAR' that lasted about 4 h, but which completely disappeared at 8 h when the DA effect had worn off. These CAR-naive rats did not learn a CAR under the influence of DA. In a third group of rats, DA produced locomotor activation which, in its time course, resembled the effect of DA on CAR. It is concluded that the ability of DA to antagonise AMPT-induced depression of CAR is, in all likelihood, dependent upon DA-induced locomotor excitation, rather than upon an effect of DA on associative learning.