Based on experimental evidence suggesting a relationship between dopamine and mania, we proposed the antidepressant-induced dopaminergic supersensitivity as a model of antidepressant-related mania. We have previously shown the ability of carbamazepine, but not lithium, to prevent this phenomenon. Here we show that sodium valproate (50 mg/kg/day for 3 weeks) fails to prevent imipramine (20 mg/kg/day for 3 weeks)-induced sensitization to the locomotor response to the dopamine D2-like receptor agonist quinpirole (0.15 mg/kg). Since lithium, carbamazepine and valproate are all considered poorly effective in the treatment of antidepressant-related mania, the validity of the proposed model should be disproved by the carbamazepine results, to which, however, a pharmacokinetic mechanism might have concurred.