Data were obtained on 255 siblicides in Canada between 1974 and 1990, 89 siblicides in England and Wales between 1977 and 1990, 45 siblicide cases in Japan that were closed in 1955, and 212 siblicides in Chicago between 1965 and 1994. Findings revealed the tendency for the killer to be the younger party was especially true of cases in which victim and killer were same-sex adults and especially brothers close in age. The older party was much more likely to be the killer when one or both were children, but this tendency was adequately accounted for by the changing age-specific likelihood that one would commit a homicide at all. Only the Japanese data set contained information on actual birth orders, which were not demonstrably related to the likelihood of either killing or being killed by a sibling. An analysis of the Canadian data suggested the rate of siblicide was not affected by the age difference between siblings. The substance of lethal sibling conflicts is discussed in light of the findings, case descriptions, and literature on non-lethal sibling conflict. 17 references, 3 tables, and 2 figures