The author contends drugs and violence may be related in three different ways: (1) the psychopharmacological model suggests some persons may become excitable and/or irrational after ingesting specific substances and may act out in a violent manner; (2) the economic compulsive model suggests some persons feel compelled to engage in economic crimes to finance costly drug use; and (3) the systemic model refers to aggressive interaction patterns within drug distribution systems. Empirical studies of the relationship between drugs and violence in New York City found relatively high proportions of violence engaged in by male and female street drug users and distributors were not related to drug use or drug trafficking. Psychopharmacological and systemic violence were the most common forms of drug-related violence. For both men and women, alcohol was the substance most likely to be associated with psychopharmacological violence. Heroin and cocaine were the substances most likely to be associated with systemic violence. Both the frequency and the volume of cocaine use were related to involvement in violence, but the nature of this involvement was quite different for men and women. Much of the violence reported by males occurred in the context of robberies or other economic crimes. The two most common types of drug-related homicide were psychopharmacological and systemic, and psychopharmacological homicides were most often alcohol-related. The relationship between drugs and violence is discussed as a public health issue. 28 references and 1 table