OBJECTIVE: The authors sought to determine whether fluphenazine dose or plasma level predicts clinical improvement or side effects during acute treatment. METHOD: Oral fluphenazine was given in fixed, randomized, double-blind doses (10, 20, or 30 mg/day) for 4 weeks to 72 inpatients with acute schizophrenic exacerbations. Outcome measures included percentage improvement in ratings of positive symptoms (hallucinations, delusions, and thought disorder), percentage improvement in negative symptoms, and maximum score for extrapyramidal symptoms. Response was defined as an improvement in positive symptoms of 40% or more. RESULTS: The 42 responders had a shorter duration of illness, less chronic course, and lower rate of akathisia. Plasma level and dose did not differentiate responders and nonresponders, but they did predict percentage improvement in positive symptoms within the responder subgroup. Akathisia was more common and extrapyramidal symptoms were more severe at higher plasma levels. CONCLUSIONS: Responders showed the greatest improvement at fluphenazine plasma levels above 1.0 ng/ml and doses above 0.20-0.25 mg/kg per day. Since the literature suggests that optimal plasma levels are similar during acute and maintenance treatment, monitoring of plasma levels may thus be useful. Conditions for applying the "responder-only" analytic strategy in future studies are discussed.Abstract Teaser