Conditioned avoidance response (CAR) behavior and catalepsy (CAT) are the standard preclinical tests used to predict antipsychotic activity and motor side-effect liability, respectively. Recent data in patients show that striatal dopamine D(2) occupancy predicts antipsychotic response (at 65% D(2) occupancy) and motor side-effects (at greater than 80%). To relate preclinical and clinical findings, this study examined the relationship between striatal D(2) occupancy, CAT and CAR in rats receiving typical and atypical antipsychotics. CAT was observed in animals receiving haloperidol, risperidone and olanzapine, but only at doses that produced a D(2) receptor occupancy > or =85%. The D(2) occupancy of quetiapine did not cross the 85% threshold (up to 100 mg/kg) and it did not show catalepsy. All drugs were effective in the CAR model at a lower level of D(2) occupancy than was required for catalepsy. We suggest that the CAR and CAT models may have displayed high predictive accuracy because they share with the clinical condition a common underlying mechanism: dopamine D(2) occupancy. The implications of this finding for understanding antipsychotic action as well as the continued use of these models in drug discovery is discussed.