The drug -violence relationship exists for several reasons, some direct (drugs pharmacologically inducing violence) and some indirect (violence occurring in order to attain drugs). Moreover, the nature of that relationship is often complex, with intoxication, neurotoxic, and withdrawal effects often being confused and/or confounded. This paper reviews the existing literature regarding the extent to which various drugs of abuse may be directly associated with heightened interpersonal violence. Alcohol is clearly the drug with the most evidence to support a direct intoxication -violence relationship. The literatures concerning benzodiazepines, opiates, psychostimulants, and phencyclidine (PCP) are idiosyncratic but suggest that personality factors may be as (or more) important than pharmacological ones. Cannabis reduces likelihood of violence during intoxication, but mounting evidence associates withdrawal with aggressivity. The literature on the relationship between steroids and aggression is largely confounded, and between 3,4-methylenedioxymethamphetamine (MDMA) and aggression insufficient to draw any reasonable conclusions. Conclusions and policy implications are briefly discussed. D