THE metabolism and tissue distribution of chlorpromazine have been studied extensively,1,2yet attempts to correlate clinical effects with variations in the fate and disposition of the drug have been unrewarding.3-7In most studies of this type the drug and its metabolites have been assayed in urine by nonspecific methods. It is becoming increasingly apparent, however, that for many drugs responses are related more closely to plasma levels of the drug (or active metabolites) than to urinary excretion rates of the drugs and their inactive metabolites.8,9Individual variation in response to drugs may be caused by variations in plasma levels, resulting from variations in rates of absorption and metabolism. Variations in plasmalevels are particularly marked in the case of highly lipid-soluble drugs, such as those used in psychopharmacology.10If these variations could be controlled by dosage regimens10,11that would