According to the dopamine hypothesis of schizophrenia, D2 receptor blockade is essential for a drug to have antipsychotic potency, and antipsychotic potency and D2 blockade are linearly related in vitro. To test this assumption in vivo, we have compared clinical response with central D2 dopamine receptor availability measured by 123I-iodobenzamide single photon emission tomography in two groups of schizophrenic patients. 6 patients were on typical antipsychotic drugs and 10 were on the atypical antipsychotic clozapine, including 2 patients from the first group. The patients on typical antipsychotics showed poor therapeutic response despite D2 receptor blockade. Significant clinical improvement occurred in all patients on clozapine, but at a lower level of D2 blockade by the drug. These findings suggest a more complex relation between D2 blockade and clinical efficacy than was previously thought.