The study hypothesized that the sociocultural etiology of homicide for infants is similar to that for adult homicide. State-by-State homicide mortality rates for infants (less than 1 year and 1-4 years) were computed from the figures on number of homicide deaths of children in these age groups as given in the 'Vital Statistics of the United States' for 1975-1980. Although the rates for males and for nonwhites were higher than for females and whites, in other respects the infant rates did not follow the adult pattern. The rates for infants did not increase between 1960 and 1980, and no significant correlations were found with 22 variables which explain a large percentage of the State-to-State variation in adult homicide. Consequently, criminal justice and public health policies based on the etiology of adult homicide may not apply to infant homicides. 1 figure, 4 tables, 21 references. (Author abstract modified)