Study data came from an examination of 502 randomly chosen cases from the files of three nonsectarian, private Boston social work agencies. The 502 cases were all those containing confirmed allegations of child abuse, child neglect, marital violence, or incest and came from a larger sample of 1,534 cases. Of the 502 cases, 50 (10 percent) had incestuous episodes. The incest relationships were mainly heterosexual, with 49 males among the 50 partners. Two-thirds of the relationships continued for several years. Victims' resistance through attempts to flee, fighting back, telling the police, and telling others was generally higher in incest cases than in other forms of family violence. Just under half of the assailants were biological fathers. Stress factors such as drunkenness, unemployment, and single parenthood tended to be less for incest perpetrators than for other perpetrators of abuse. Incest is a social problem largely because it is coercive. It does not appear to be a tabooed behavior because of its frequency and the absence of strong prohibitive conditioning. Finally, it does not appear to be provoked by severe external stress. Further research should focus on internal family dynamics and power relations, because the pattern most consistently associated with incest was extreme male domination of the family. Data tables, notes, and 28 references are supplied.