THE STUDY MATCHED 20 MOTHERS AGAINST WHOM CHILD ABUSE REPORTS HAD BEEN FILED WITH 20 MOTHERS WHO BROUGHT THEIR CHILDREN TO CHILDREN'S HOSPITAL OF MICHIGAN FOR EMERGENCY ILLNESSES. THIS CONTROL GROUP WAS SELECTED BECAUSE THE STRESS PLACED ON A CHILD AND FAMILY IS SIMILAR FOR BOTH TYPES OF HOSPITAL ADMISSION. PSYCHOLOGICAL TESTS, ATTITUDE QUESTIONNAIRES, AND PERSONNEL VISITS WERE USED TO INVESTIGATE FACTORS ASSOCIATED WITH ABUSE AS REPORTED IN EXISTING LITERATURE. THE STUDY WAS PRESENTED TO THE MOTHERS AS A GENERAL SURVEY OF CHILDHOOD TRAUMA. (THE RECRUITMENT LETTER AND SURVEY INSTRUMENTS ARE INCLUDED, AS ARE TABLES PRESENTING STUDY FINDINGS.) IN GENERAL IT WAS FOUND THAT ABUSED CHILDREN ARE YOUNG (MOST UNDER 3 YEARS), THAT THEIR INJURIES ARE MULTIPLE, AND THAT DEATH OR SERIOUS DISABILITY OFTEN RESULTS. MATERNAL FACTORS ARE FOUND TO BE THE MOST IMPORTANT PREDICTORS. ABUSIVE MOTHERS ARE MORE ALIENATED AND SOCIALLY NONCONFORMING. WHEN SOCIALLY ACCEPTED ANSWERS ON A QUESTIONNAIRE ARE OBVIOUS, THEY GIVE SOCIALLY ACCEPTED ANSWERS. HOWEVER, LESS OBVIOUS QUESTIONS SHOW THAT THESE MOTHERS HAVE UNREALISTIC EXPECTATIONS FOR THEIR CHILDREN. IT ALSO FOUND THAT SIBLINGS WERE FREQUENTLY ABUSED. INVESTIGATORS ARE WARNED AGAINST THE MYTH THAT ABUSE FOCUSES ON ONLY ONE CHILD. AS PREVIOUSLY REPORTED IN THE LITERATURE, ABUSING MOTHERS HAD LOWER INTELLIGENCE AND FEWER SOCIAL RESOURCES. CONTRARY TO PREVIOUS REPORTS, TREATMENT AS A CHILD WAS NOT SIGNIFICANT. AS MANY NONABUSING AS ABUSING MOTHERS REPORTED HARSH PHYSICAL PUNISHMENT. CHILD-RELATED FACTORS DID NOT SEEM SIGNIFICANT IN ABUSE PREDICTION. THE SURVEY FINDINGS ARE DISCUSSED IN DETAIL. REFERENCES ARE INCLUDED. (GLR)