Background: Thioridazine is an antipsychotic that can still be used for schizophrenia although it is associated with the cardiac arrhythmia, torsades de pointe. Objectives: To review the effects of thioridazine for people with schizophrenia. Search strategy: For this 2006 update, we searched the Cochrane Schizophrenia Group's Register (June 2006). Selection criteria: We included all randomised clinical trials comparing thioridazinewith other treatments for peoplewith schizophrenia or other psychoses. Data collection and analysis: We reliably selected, quality rated and extracted data from relevant studies. For dichotomous data, we estimated relative risks (RR), with the 95% confidence intervals (CI). Where possible, we calculated the number needed to treat/harm statistic (NNT/H) on an intention-to-treat basis. Main results: This review currently includes 42 RCTs with 3498 participants. When thioridazine was compared with placebo (total n=668, 14 RCTs) we found global state outcomes favoured thioridazine (n=105, 3 RCTs, RR 'no change or worse' by 6 months 0.33 CI 0.2 to 0.5, NNT of 2 CI 2 to 3). Thioridazine is sedating (n=324, 3 RCTs, RR 5.37 CI 3.2 to 9.1, NNH 4 CI 2 to 74). Generally, thioridazine did not cause more movement disorders than placebo. Twenty-seven studies (total n=2598) compared thioridazine with typical antipsychotics. We found no significant difference in global state (n=743, 11 RCTs, RR no short-term change or worse 0.98 CI 0.8 to 1.2) and medium-term assessments (n=142, 3 RCTs, RR 0.99, CI 0.6 to 1.6). We found no significant differences in the number of people leaving the study early 'for any reason' (short-term, n=1587, 19 RCTs, RR 1.07 CI 0.9 to 1.3). Extrapyramidal adverse events lower for those allocated to thioridazine (n=1082, 7 RCTs,. RR use of antiparkinsonian drugs 0.45 CI 0.4 to 0.6). Thioridazine did seem associated with cardiac adverse effects (n=74, 1 RCT, RR 'any cardiovascular adverse event' 3.17 CI 1.4 to 7.0, NNH 3 CI 2 to 5). Electrocardiogram changes were significantly more frequent in the thioridazine group (n=254, 2 RCTs, RR 2.38, CI 1.6 to 3.6, NNH 4 CI 3 to 10). Six RCTs (total n=344) randomised thioridazine against atypical antipsychotics. Global state rating did not reveal any short-term difference between thioridazine and remoxipride and sulpiride (n=203, RR not improved or worse 1.00 CI 0.8 to 1.3). Limited data did not highlight differences in adverse event profiles. Authors' conclusions: Although there are shortcomings, there appears to be enough consistency over different outcomes and periods to confirm that thioridazine is an antipsychotic of similar efficacy to other commonly used antipsychotics for people with schizophrenia. Its adverse events profile is similar to that of other drugs, but it may have a lower level of extrapyramidal problems and higher level of ECG changes. We would advocate the use of alternative drugs, but if its use in unavoidable, cardiac monitoring is justified.