The children all lived in upstate New York. The children in foster care included 25 who were placed with their siblings and 13 who were separated from their siblings. Data were collected from the foster mothers and the mothers of those living at home. The data were analyzed by means of chi-square, ANOVA, and paired t-tests that examined differences between groups and within participants in the children's friendships. Results revealed that the majority of the high-risk children had a best friend. However, the foster children all had friends who could be described as quasi-siblings. Findings suggest that foster care placement and the presence of siblings may influence the nature of a child's interactions with unrelated peers; those separated from their siblings directed more negative behavior toward their friends than the others and those placed with their siblings directed more positive behavior toward their friends than did comparison children living at home with siblings. Findings indicated that the maltreated children in the sample seemed capable of forming and maintaining some form of friendship in home and neighborhood settings. Tables and 52 references (Author abstract modified)