To compare the remission rate and its time-course over 196 wk of double-blind treatment with an atypical antipsychotic, ziprasidone (80-160 mg/d given b.i.d., or 80-120 mg/d given q.d.), or a conventional antipsychotic, haloperidol (5-20 mg/d). Outcome assessments included attainment of remission (Andreasen criteria) by longitudinal analysis. Positive and Negative Syndrome Scale (PANSS) scores, Global Assessment of Functioning Scale (GAF) scores, and quality-of-life (QLS) were also assessed in the initial 40-wk study phase (n=599) and the 3-yr extension study (n=186). Discontinuation rates in the initial 40-wk core and follow-up extension studies were comparable between groups: 64% and 65% for the 80-160 mg/d ziprasidone group, 65% and 58% for the 80-120 mg/d ziprasidone group, and 60% and 66% for the 5-20 mg/d haloperidol group, respectively. Mean change scores from baseline to LOCF endpoint (week 40 or early termination) for PANSS negative and GAF (primary efficacy variables) were not statistically significantly different between ziprasidone and haloperidol. During the 3-yr extension study, ziprasidone-treated subjects (80-160 mg/d) were more likely to achieve remission (51%) than haloperidol-treated (40%) subjects (p=0.04), while there was a favourable trend associated with 80-120 mg/d ziprasidone (48%). Compared to the haloperidol group, subjects assigned to the 80-160 mg/d ziprasidone group showed a gradual and persistent improvement in remission (p=0.006) and quality-of-life (p=0.004) in the longitudinal analyses. Significant differences in the trajectory of PANSS total and GAF scores favouring the 80-160 mg/d ziprasidone group were also observed. In this long-term, double-blind study, ziprasidone treatment was more likely to result in remission than haloperidol treatment, and was associated with greater improvement in quality-of-life.