Sex- and age-specific rates of killing unrelated persons of one’s own sex were computed for Canada (1974–1983), England/Wales (1977–1986), Chicago (1965–1981), and Detroit (1972) from census information and data archives of all homicides known to police. Patterns in relation to sex and age were virtually identical among the four samples, although the rates varied enormously (from 3.7 per million citizens per annum in England/Wales to 216.3 in Detroit). Men’s marital status was related to the probability of committing a same-sex, nonrelative (...) homicide, but age effects remained conspicuous when married and unmarried men were distinguished.These findings and the treatment of age and sex effects by criminologists are discussed in the light of contemporary evolutionary psychological models of sex differences and life-span development. Same-sex homicides in which killer and victim are unrelated can be interpreted as an assay of competitive conflict. In every human society for which relevant information exists, men kill one another vastly more often than do women. Lethal interpersonal competition is especially prevalent among young men, which accords with many other aspects of life-span development in suggesting that sexual selection has maximized male competitive prowess and inclination in young adulthood. (shrink)