Aggressive behavior of psychotic patients impacts all aspects of their clinical care. Better treatments to address this problem are needed, and atypical antipsychotics, such as clozapine, risperidone, and perhaps quetiapine, have shown promise. However, studying the psychopharmacology of aggression is difficult because of the many methodological problems that arise in the design of appropriate clinical trials. These include imprecise definitions of aggression, the difficulty of measuring outcome because of the relative rarity of aggressive events, bias in the selection of patients for study, inadequate and inappropriate control groups, and inattention to comorbidities and concomitant medications in analyzing results. Since the usual outcome measure is the aggressive event rate, a large sample size and lengthy baseline and trial periods are required when this rate is low. Furthermore, formidable practical and ethical obstacles interfere with the many sound techniques (e.g. randomization) used in typical designs of psychopharmacological clinical trials. Current research methods should be modified and new ones developed in order to progress in assessing the antiaggressive effects of treatments.