Support for 'legitimate violence' is reflected in an underlying cultural belief in the efficacy and desirability of physical force. In this study, legitimate violence was measured by an index combining 12 indicators of noncriminal violence (e.g., enrollment rate for the National Guard and the circulation rate of violent magazines). The second index for measuring legitimate violence, the Violence Approval Index, was computed by combining responses to 14 questions on situations in which the respondents approved of using violence. A theoretical model hypothesizing the relationship of these two measures of cultural support for violence (and seven control variables) to rape was developed and tested using path analysis. Findings show that legitimate violence is directly related to the rape rate. The degree of social disorganization, urbanization, economic inequality, and percent of single males are also directly related to rape. The youthfulness of the population and the percentage of blacks affect rape indirectly through their association with legitimate violence. These findings support structural explanations of the origins of cultural support for violence and for a 'cultural spillover' theory of rape. 5 tables, 4 figures, 55 references, 12 footnotes. (Author abstract modified)