Aggressive behavior complicates the presentation of many psychiatric illnesses, and is associated with significant morbidity. Antipsychotic medications are used to treat this symptom dimension across multiple diagnoses. In this meta-analysis we sought to identify the effect size of antipsychotic medications for the treatment of reactive-impulsive aggression in adults, and identify differences across underlying diagnosis and specific agent. A search was conducted of four databases, MEDLINE, PsychINFO, Embase and the Cochrane Library to end date of August 10, 2016. The search terms included “aggression”, “irritable mood”, “anger”, “hostility” and “antipsychotic agents” or “dopamine antagonists”. 505 results were found, of which 47 were reviewed in detail and 21 ultimately included in the analysis. Antipsychotics were broadly effective for the treatment of aggression, but with effect sizes similar to those for non-pharmacologic interventions (standard mean difference = 0.29, 95% confidence interval 0.22–0.36, z = 8.5, p < 0.001). There was no evidence for differences according to choice of agent (χ2 = 2.7, df = 6, p = 0.85), or conclusive evidence as to the importance of the underlying diagnosis (χ2 = 3.2, df = 3, p = 0.36). A small but significant dose effect was identified (β = 0.0002, 95% CI 0.0001–0.0004, p = 0.038). Although antipsychotics appear to be effective for treatment of aggression, their small effect sizes in the context of their significant side-effects should be taken into account when making clinical decisions about their use.