Attempts to investigate relationships between plasma levels of neuroleptics and therapeutic outcome in schizophrenic patients have been hampered due to such factors as the chemical nature of these drugs, their metabolism, and the very heterogeneous nature of the disease states. Two clinical studies are described that investigate the relationship between plasma levels of fluphenazine (FLU) and its metabolites and therapeutic outcome in schizophrenic patients. In the first of these studies the levels of FLU and fluphenazine sulfoxide (FLUSO) in schizophrenics receiving either 5 or 25 mg of fluphenazine decanoate (FLUD) intramuscularly every 2 weeks were monitored. Patients given 25 mg of FLUD required 3 months to reach plasma level steady state. The results suggest that such patients, when being switched from the oral to the depot formulation of FLU, should continue to receive oral supplementation during the 1st 3 months after conversion. The relationship between log-transformed plasma levels at 26 and 38 weeks with subsequent psychotic exacerbation was investigated with the use of logistic regression and survival analysis. Both demonstrated significant relationships between FLU plasma levels and a risk of psychotic exacerbation at 26 and 38 weeks. The possibility of any correlations between neurological side effects and plasma concentrations were also investigated, with statistically significant correlations between FLU levels and akinesia found at 2 and 26 weeks. In the second of these studies the levels of FLU, FLUSO, 7-hydroxyfluphenazine (7-OHFLU), and fluphenazine N4‘-oxide (FLUNO) in schizophrenics receiving 5, 10, or 20 mg of oral fluphenazine dihydrochloride daily for 4 weeks were monitored. The relationships between log-transformed plasma levels, disabling side effects, and global improvement were examined by logistic regression for the 4-week period. The study showed a significant correlation between increases in both plasma levels and disabling side effects such that at a plasma level of 2.7 ng/ml, approximately 90% of acutely ill patients experienced disabling side effects. Conversely, the study also showed that at a plasma level of 0.67 ng/ml, 48% of patients experienced improvement without the development of disabling side effects. When relationships between metabolite levels, disabling side effects, and global improvement were examined by logistic regression, a stronger correlation between disabling side effects and FLUNO levels than between side effects and FLU levels was found. No correlations between disabling side effects and plasma levels of FLUSO or 7-OHFLU were observed. These studies illustrate the usefulness of measuring plasma levels of neuroleptic drugs.