The outcome of experiencing abuse as a child was studied by comparing pregnant women who did recall versus those who did not recall this type of maltreatment. Fourteen hundred low income women were interviewed in prenatal clinic: those who said they were both punished by abuse and beaten by caretakers as children were considered abused. Abuse was recalled by more white than black women but families were followed after delivery and protective service reports of abuse for their offspring were equal. Because of low numbers, black subjects were dropped and the 95 white women who recalled abuse during childhood were compared to the remaining 832 while subjects. The groups did not differ in attitude about current pregnancy, age or marital status, and no differences were found for their children at birth. Abused mothers were more likely to have felt unwanted and unloved as children and to have lower self-images and more isolation than controls. Abused mothers had greater stress, and many of their stresses reflected disturbances in interpersonal relationships. Thus, women abused as children had some characteristics similar to those of known child abusers. Although abused women had more aggressive tendencies, their children were reported to protective services for abuse at the same frequency as control children. Intergenerational transmission of abuse was therefore not demonstrated prospectively. Classic theories of child abuse suggest a special child, special parent and stress act as independent agents to cause abuse. The above data suggest, alternatively, that abuse during childhood may lead to other risk characteristics and to greater stress. These may act together to increase risk for abuse. The special child may have an independent influence on abuse.