The descriptive data for 1,202 alleged perpetrators show significant differences between males and females. Female perpetrators were more likely to be young, single, poor Hispanics; while male perpetrators were more likely to be older, married, middle-class Anglos or blacks. Multiple regression analysis suggests differences in the nature of social support that appear to mediate maltreatment by males and females. Neighborhood maltreatment rates by females were lower when substantial numbers of women with access to material resources were available for support. Female rates were higher in neighborhoods with a disruption of social networks caused by the arrival of large numbers of new residents. Neighborhood rates of maltreatment by males appear to be strongly influenced upward by large numbers of women in the neighborhood who are single parents and, therefore, less available to give support. Limitations on material resources associated with unemployment in the neighborhood also appears to contribute substantially to male maltreatment rates. Results suggest that children in neighborhoods socially impoverished because of structural inequalities are at risk of maltreatment. 2 tables, 6 notes, and 26 references. (Author abstract modified)