While dopamine D2 receptor partial agonists (PAs) have been long considered for treating schizophrenia, only one, aripiprazole, is clinically available for therapeutic use. This raises critically important questions as to what is unique about aripiprazole and to what extent animal models can predict therapeutic success. A number of PAs whose clinical fate is known: aripiprazole, preclamol, terguride, OPC-4392 and bifeprunox were compared to haloperidol (a reference antipsychotic) in several convergent preclinical animal models; i.e. amphetamine-induced locomotion (AIL) and conditioned avoidance response (CAR), predictive of antipsychotic effects; unilateral nigrostriatal lesioned rats, a model of hypo-dopaminergia; striatal Fos induction, a molecular marker for antipsychotic activity; and side-effects common to this class of drugs: catalepsy (motor side-effects) and prolactaemia. The results were compared across drugs with reference to their measured striatal D2 receptor occupancy. All the PAs occupied striatal D2 receptors in a dose dependent manner, inhibited AIL and CAR, and lacked motor side-effects or prolactinaemia despite D2 receptor occupancy exceeding 80%. At comparative doses, aripiprazole distinguished itself from the other PAs by causing the least rotation in the hypo-dopaminergic model (indicating the least intrinsic activity) and showed the highest Fos expression in the nucleus accumbens (indicating functional D2 antagonism). Although a number of PAs are active in antipsychotic animal models, not all of them succeed. Given that only aripiprazole is clinically available, it can be inferred that low functional intrinsic activity coupled with sufficient functional antagonism as reflected in the animal models may be a marker of success.