PURPOSE: Atypical antipsychotic agents are used to treat the symptoms of schizophrenia and bipolar disorder. The purpose of this review is to help policy makers and clinicians make informed choices about their use. Given the prominent role of drug therapy in psychiatric disease, our goal is to summarize comparative data on efficacy, effectiveness, tolerability, and safety. Ten atypical antipsychotics are currently available in the United States and Canada. Clozapine, the prototypic atypical antipsychotic, was introduced in 1989. Since then, 9 other atypical antipsychotics have been brought to market: risperidone (1993), risperidone long-acting injection (2003), olanzapine (1996), quetiapine (1997), ziprasidone (2001), aripiprazole (2002), extended-release paliperidone (2006), asenapine (2009), iloperidone (2009), and paliperidone long-acting injection (2009). DATA SOURCES: To identify relevant citations, we searched the Cochrane Central Register of Controlled Trials (1st Quarter 2010), Cochrane Database of Systematic Reviews (4 quarter 2009), MEDLINE (1950 to week 4 January 2010), and PsycINFO (1806 to February week 1 2010) using terms for included drugs, indications, and study designs. We attempted to identify additional studies through searches of reference lists of included studies and reviews. We also searched the US Food and Drug Administration Center for Drug Evaluation and Research website for medical and statistical reviews of individual drug products. Finally, we requested dossiers of published and unpublished information from the relevant pharmaceutical companies for this review. REVIEW METHODS: Study selection, data abstraction, validity assessment, grading the strength of the evidence, and data synthesis were all carried out according to our standard review methods. RESULTS: In patients with schizophrenia, while differences in short-term efficacy are not apparent among the atypical antipsychotics, clozapine and olanzapine have been found to result in lower rates of discontinuation of drug over periods of up to 2 years. Clozapine has reduced suicides and suicidal behavior in patients at high risk, but results in more discontinuations due to adverse events than the others. While risperidone and extended-release paliperidone resulted in higher rates of extrapyramidal symptoms in some studies, the majority of studies find no differences among the drugs. Risperidone was found to result in more frequent or more severe sexual dysfunction symptoms than quetiapine, but was similar to extended-release paliperidone or ziprasidone. Very limited evidence existed regarding atypical antipsychotics used for the treatment of schizophrenia in subgroup populations. Among adolescents with schizophrenia, quetiapine was not superior to placebo based on response rate, but was superior based on improvements measured by the Positive and Negative Syndrome Scale. Differences by race were not found, but women had greater improvements with clozapine on a global impression scale, and with olanzapine on a quality of life scale compared with men. In adults with bipolar disorder, no significant differences were found between risperidone and olanzapine or asenapine and olanzapine in quality of life, remission, and response outcomes. Olanzapine resulted in greater mean weight gain compared with asenapine and risperidone, respectively, whereas asenapine resulted in a significantly higher rate of discontinuations due to adverse events than olanzapine. Otherwise, there were no significant differences between risperidone and olanzapine or between asenapine and olanzapine in extrapyramidal symptoms or between risperidone and olanzapine in discontinuations due to adverse events. In children and adolescents with bipolar disorder evidence is extremely limited; olanzapine and risperidone had similar response rates after 8 weeks of treatment and no significant differences in mean weight gain were found. In adults with major depressive disorder, the majority of studies evaluated the adjunctive use of atypical antipsychotics in patients with an inadequate response to prior treatment with standard antidepressants and generally provided insufficient evidence for determining their comparative effectiveness and efficacy. However, evidence from both observational studies and randomized controlled trials indicated that weight gain was greatest with adjunctive olanzapine. In patients with behavioral and psychological symptoms of dementia, the best evidence found similar rates of response and withdrawal, and no differences in clinical outcome measures for olanzapine, risperidone, and quetiapine. Compared with placebo, risperidone, aripiprazole, and olanzapine improved behavioral symptoms in children and adolescents with pervasive developmental disorders, and risperidone and quetiapine showed efficacy in children and adolescents with disruptive behavior disorders. Olanzapine resulted in greater weight gain compared with other atypical antipsychotics (6 to 13 pounds more), and an increased risk of new-onset diabetes (OR, 1.16; 95% CI, 1.0 to 1.31 compared with risperidone). Risperidone resulted in an increased risk of new-onset tardive dyskinesia (3% compared with 1% to 2% for others). While clozapine has been shown to be associated with increased risk of seizures and agranulocytosis, differences among the drugs in other serious harms have not been clearly shown. CONCLUSION: Few differences were seen among the atypical antipsychotics in short-term efficacy in patients with schizophrenia, bipolar disorder, or dementia. Differences in most effectiveness outcomes were also not clear, but uncertainty exists. In patients with schizophrenia, clozapine reduced suicides and suicidal behavior, but resulted in stopping drug due to adverse events more often than the others. However, clozapine and olanzapine resulted in lower rates of discontinuation of drug for any reason over periods of up to 2 years. In adults with bipolar disorder, asenapine resulted in a higher risk of stopping drug due to adverse events than olanzapine. Comparative evidence was not available for the use of the drugs in adults with major depressive disorder or children and adolescents with pervasive developmental disorders or disruptive behavior disorders. Olanzapine resulted in greater weight gain than the other drugs (6 to 13 pounds more) and a 16% increased risk of new-onset diabetes, while risperidone resulted in an increased risk of new-onset tardive dyskinesia. While clozapine has been shown to be associated with increased risk of seizures and agranulocytosis, differences among the drugs in other serious harms have not been clearly shown. Evidence on long-term harms for the newest drugs is lacking.