A review of these findings revealed literature in the early stages of development. Nothing is known about the ecology of sexual abuse, and studies of physical abuse and neglect have done little more than demonstrate co-variation between reported incidence and neighborhood population and housing characteristics. Study findings reveal that (a) five of the seven community characteristics are significant and independent correlates of neglect, and four are correlates of abuse; (b) the strongest predictors of both types of maltreatment are percentage of families with income less than 200% of poverty and percent vacant housing; and (c) the pattern of co-variation between the two economic stress indicators and three inadequate social support indicators are consistent with the ecological hypothesis. Discussion focuses on interpretation of findings and recommendations for future research. 3 tables and 49 references. (Author abstract)