Because of large interindividual variabilities in the pharmacokinetics of haloperidol (HPDL), empirically adjusting the dose to achieve steady-state levels in plasma (Css) is a time-consuming process. We report a method to individualize dose to achieve a desired Css from an observed drug level 24 hours after a single 15-mg test dose of HPDL. Drug-free schizophrenic and schizo-affective patients were blindly and randomly assigned to achieve a low (< 5 ng/ml), medium (10-18 ng/ml), or high (> 25 ng/ml) Css range of HPDL. On day 1 of the study, each patient received an oral "test" dose of HPDL (15 mg), and blood was drawn 24 hours later to determine drug levels in plasma (C24h). The first 34 patients (group I) were then maintained empirically on 2, 5 to 8, or 10 to 15 mg twice daily of oral HPDL concentrate for 5 days to achieve a low, medium, or high Css range, respectively. For the next 58 patients (group II), the dose of HPDL to achieve the assigned Css range was computed by using C24h in a prediction formula. Application of the C24h correctly predicted the maintenance dose required to achieve the Css in 73.2% of the cases. Further, there was a highly significant correlation (R2 = 0.877, p < 0.0001) between the predicted dose and the actual dose required to achieve the targeted Css range. On the basis of these results, we have formulated a nomogram to help predict the maintenance dose required to achieve low, medium, or high HPDL targeted ranges at various C24h values.