Reducing the risk of violent and aggressive behaviour in patients with schizophrenia remains a clinical priority. There is emerging evidence to suggest that the second-generation antipsychotic, clozapine, is effective at reducing this risk in patients with schizophrenia and some evidence to suggest that it may be best in selected patients. We conducted a systematic literature search in March 2011 of all prospective and retrospective studies, which investigated clozapine's anti-aggressive effects in a variety of mental disorders. The review identified six animal studies, four randomized controlled trials, 12 prospective non-controlled studies and 22 retrospective studies, with four case studies. We found considerable evidence in support of clozapine's ability to reduce violent and aggressive behaviour. Clozapine's anti-aggressive effect was most commonly explored in patients with schizophrenia, with less evidence available for other psychiatric disorders, including borderline personality disorder, autistic spectrum disorders, post-traumatic stress disorder, bipolar disorder and learning disability. There was mixed evidence to address the question of whether or not clozapine was any more effective than other antipsychotics. In the case of schizophrenia, there was evidence to suggest that clozapine's anti-aggressive effect was more marked particularly in those with treatment-resistant illness. Its anti-aggressive effects appeared to be ‘specific’, being to some extent greater than both its more general antipsychotic and sedative effects. There were significant methodological inconsistencies in the studies we identified, particularly surrounding patient recruitment criteria, the definition and measurement of violence and the lack of randomized, controlled trials. Data on therapeutic monitoring were also limited. Clozapine can reduce violence and persistent aggression in patients with schizophrenia and other psychiatric disorders. It may offer an advantage over other antipsychotics, although perhaps exclusively in the case of traditionally defined ‘treatment resistance’ or more broadly defined ‘complex cases’ with co-morbidity. Larger, randomized, blinded, controlled studies with robust characterization of participants, and standardized measures of violence and aggression are, however, needed to fully understand this link and explore the possible mechanisms.