Amisulpride is an atypical antipsychotic with selective affinity for dopamine D2/3 receptors. In this long-term, open, randomised, multicentre trial, patients with chronic or subchronic schizophrenia received amisulpride (n =370) or haloperidol (n = 118) for 12 months. Dosage regimens were flexible (amisulpride 200-800 mg/day, haloperidol 5-20 mg/day). Improvement in mean Brief Psychiatric Rating Scale total score was significantly greater for amisulpride than haloperidol (17.0 versus 12.8, P = 0.01). Positive symptoms (Positive and Negative Syndrome Scale [PANSS] positive) improved in a similar way in each group but amisulpride caused a significantly better improvement in negative symptoms (PANSS negative) (7.1 versus 3.7, P < 0.0001). Improvements in Global Assessment of Functioning (GAF) and Quality of Life Scale (QLS) scores were also significantly greater in the amisulpride group (GAF -20.1 versus -13.6, P = 0.001; QLS -0.64 versus -0.30, P = 0.02). Adverse events were mainly psychiatric in nature, and occurred with similar frequency in each group (amisulpride 254/370, 69%; haloperidol 82/118, 70%). Extrapyramidal symptoms were more frequent for haloperidol (48/118, 41% versus 96/370, 26% for amisulpride), leading to a greater requirement for antiparkinsonian medication (haloperidol 66/118, 56% versus amisulpride 118/370, 32%). Haloperidol significantly aggravated parkinsonism, akathisia and involuntary movement compared to amisulpride. The overall incidence of endocrine events was comparable between groups (4% for amisulpride, 3% for haloperidol). Maintenance of efficacy was comparable in both treatment groups; 59% of amisulpride patients and 55% of haloperidol patients improved after 1 month of therapy remained improved throughout the study period. Amisulpride is effective following flexible long-term administration and significantly improves social functioning and quality of life.