RATIONALE: l-Stepholidine, a dopamine D(2) antagonist with D(1) agonist activity, should in theory control psychosis and treat cognitive symptoms by enhancing cortical dopamine transmission. Though several articles describe its impact on the dopamine system, it has not been systematically evaluated and compared to available antipsychotics. MATERIALS AND METHODS: We examined its in vitro interaction with dopamine D(2) and D(1) receptors and compared its in vivo pharmacokinetic profile to haloperidol (typical) and clozapine (atypical) in animal models predictive of antipsychotic activity. RESULTS: In vitro, l-stepholidine showed significant activity on dopamine receptors, and in vivo, l-stepholidine demonstrated a dose-dependent striatal receptor occupancy (RO) at D(1) and D(2) receptors (D(1) 9-77%, 0.3-30 mg/kg; D(2) 44-94%, 1-30 mg/kg), though it showed a rather rapid decline of D(2) occupancy related to its quick elimination. In tests of antipsychotic efficacy, it was effective in reducing amphetamine- and phencyclidine-induced locomotion as well as conditioned avoidance response, whereas catalepsy and prolactin elevation, the main side effects, appeared only at high D(2)RO (>80%). This preferential therapeutic profile was supported by a preferential immediate early gene (Fos) induction in the nucleus accumbens over dorsolateral striatum. We confirmed its D(1) agonism in vitro, and then using D(2) receptor, knockout mice showed that l-stepholidine shows D(1) agonism in the therapeutic dose range. CONCLUSIONS: Thus, l-stepholidine shows efficacy like an "atypical" antipsychotic in traditional animal models predictive of antipsychotic activity and shows in vitro and in vivo D(1) agonism, and, if its rapid elimination does not limit its actions, it could provide a unique therapeutic approach to schizophrenia.