BACKGROUND: Acute psychotic illness, especially when associated with agitated or violent behaviour, can require urgent pharmacological tranquillisation or sedation. In several countries, clinicians often use benzodiazepines (either alone or in combination with antipsychotics) for this outcome. OBJECTIVES: To estimate the effects of benzodiazepines, alone or in combination with antipsychotics, when compared to placebo or antipsychotics, to control disturbed behaviour and reduce psychotic symptoms. SEARCH STRATEGY: We searched the Cochrane Schizophrenia Group's register (October 2002 and April 2005), inspected reference lists of included and excluded studies and contacted authors of relevant studies. SELECTION CRITERIA: We included all randomised clinical trials comparing benzodiazepines, alone or in combination with antipsychotics, with placebo or sole use of antipsychotics, for people with acute psychotic illnesses. DATA COLLECTION AND ANALYSIS: We reliably selected studies, quality assessed them and extracted data. For binary outcomes we calculated standard estimates of relative risk (RR) and their 95% confidence intervals (CI), and weighted number needed to treat or harm (NNT/NNH) statistics. For continuous outcomes we estimated a weighted mean difference between groups. If heterogeneity was found, we used a random effects model. MAIN RESULTS: We included eleven studies with a total of 648 participants. When comparing benzodiazepines with placebo, sedation was equally prevalent (n=102, 1 RCT, RR 1.67 CI 0.4 to 6.6), however, fewer people allocated lorazepam remained excited at 24 hours (n=102, RR 0.62 CI 0.4 to 1.0, NNT 5 CI 3 to 59). The lorazepam and placebo group experienced similar non-significant, low levels of adverse effects. In the comparison of benzodiazepines versus use of antipsychotics without use of anticholinergics/antihistamines, people allocated benzodiazepines did not clearly need additional medication compared with those given antipsychotics (n=216, 2 RCTs, RR 1.28 CI 0.5 to 3.2). Numbers sedated were also equivocal between groups (n=324, 6 RCTs, RR 0.76 CI 0.5 to 1.2) as were mental state ratings. Extrapyramidal symptoms were significantly higher in the antipsychotic treatment group (n=391, 7 RCTs, RR 0.17 CI 0.1 to 0.4, NNT 6 CI 2 to 17). Two trials (total n=83) comparing lorazepam plus haloperidol with lorazepam alone found no clear difference for the need of additional medication (n=83, 2 RCTs, RR 1.02 CI 0.8 to 1.3) or 'not improved' at one hour (n=20, 1 RCT, RR 1.47 CI 0.66 to 3.25). There was no difference in the incidence of extrapyramidal symptoms (n=83, 2 RCTs, RR 1.94 CI 0.2 to 20.3). Finally when the benzodiazepine plus antipsychotic combination was compared with antipsychotics alone (2 RCTs, n=95) there was no difference between groups in the need for additional medications (n=67, 1 RCT, RR 0.95 CI 0.8 to 1.2) or for mental state measures. Extrapyramidal symptoms were significantly lower for people receiving both benzodiazepines and antipsychotics compared with those receiving antipsychotics alone (n=95, 2 RCTs, RR 0.45 CI 0.2 to 0.9, NNH 2 CI 1 to 5). There was no significant difference in the number of participants unfit for early discharge (n=28, 1 RCT, RR 0.90 CI 0.54 to 1.5). AUTHORS' CONCLUSIONS: There is insufficient data from these studies to support or refute the use of benzodiazepines with or without antipsychotics where emergency drugs are needed. The sole use of older antipsychotics unaccompanied by anticholinergic drugs may be problematic, but studies in this review are not large enough to identify any serious adverse effects of benzodiazepines such as respiratory depression. Larger, more informative studies are needed before definite conclusions can be drawn as to the efficacy of benzodiazapines.