Prior event-based research regarding the relationship between alcohol and violence suffers from important limitations, including the absence of a comparison group, an inappropriate comparison group, or a comparison group that could be considered appropriate but does not control for potential confounders. To overcome such limitations, we use a matched pair design. Drawing on interviews with men imprisoned for an aggravated assault or homicide committed in response to a conflict with another man, we examine matched pairs of violent and nonviolent conflicts nested within respondents. The results suggest that guns mediate the relationship between alcohol and lethal male-male violence. (Contains 4 tables and 11 notes.)