SUBJECTS WERE INNER-CITY RESIDENTS REFERRED TO A FORENSIC PSYCHIATRY DIVISION BETWEEN JANUARY 1974 AND DECEMBER 1975, WHO COMPLETED THE REQUISITE EVALUATIVE STUDIES AND ASSESSMENTS. THE SAMPLE WAS 86 PERCENT FEMALE AND 14 PERCENT MALE. THE MEAN AGE WAS 25.33 YEARS, MEAN EDUCATIONAL LEVEL 10.94 YEARS, AND MEAN INTELLIGENCE QUOTIENT WAS 91.30; 95 PERCENT OF THE POPULATION WAS BLACK. DESPITE PROBLEMS OF WORKING WITH A COURT-REFERRED CLINICAL SAMPLE, SUCH AS SMALL SIZE, POSSIBLE BIAS RESULTING FROM A HIGH NO-SHOW RATE, AND THE LACK OF CONTROLS, A NUMBER OF SAMPLE-SPECIFIC GENERALIZATIONS ABOUT CHILD ABUSE AND ABUSER FACTORS CAN BE MADE. CONTRARY TO THE ASSUMPTION THAT CHILD ABUSE IS A LEARNED BEHAVIOR WITH AN INTERGENERATIONAL PATTERN, OVER HALF OF THESE ABUSERS REPORTED THAT THEY HAD NOT BEEN ABUSED AS CHILDREN, ALTHOUGH ALMOST HALF HAD UNSTABLE FAMILY BACKGROUNDS AND A MAJORIY CAME FROM LARGE FAMILIES. THIS SUGGESTS THAT EMOTIONAL DEPRIVATION MAY BE A FACTOR IN CHILD ABUSE. THE ASSUMPTION OF SOCIAL ISOLATION OF CHILD ABUSING PARENTS WAS SUPPORTED IN ALL ASPECTS OF THESE PARENTS' LIVES. ALMOST ALL ABUSERS REPORTED A LACK OF CLOSE AND SATISFACTORY INTERPERSONAL RELATIONSHIPS. WHILE THE LITERATURE SUGGESTS THAT MOST ABUSERS WILL DENY ABUSE, AS MANY ABUSERS ADMITTED THE EVENT AS DENIED IT IN THIS SAMPLE. HOWEVER, ONLY 16 PERCENT ADMITTED PERSONAL RESPONSIBILITY FOR THE BATTERING EPISODE. AGAIN CONTRARY TO EXPECTATION, THE INCIDENCE OF PSYCHOPATHOLOGY IN THIS SAMPLE WAS QUITE LOW, AS WAS THE NUMBER OF BORDERLINE MENTAL DEFECTIVES. A MAJORITY OF THE SAMPLE DID EXHIBIT PERSONALITY DISORDERS INCLUDING IMPULSIVITY, CHRONIC ANGER, AND DEPENDENCY. FINDINGS PERTINENT TO CHILD CHARACTERISTICS OFTEN ASSOCIATED WITH CHILD ABUSE WERE AGAIN CONTRARY TO GENERAL ASSUMPTIONS; FOR EXAMPLE, NO CHILD OF A PARTICULAR AGE GROUP WAS MORE FREQUENTLY ABUSED; THE PROBLEM CHILD (PHYSICAL, MENTAL, OR EMOTIONALLY HANDICAPPED) WAS NOT MORE FREQUENTLY ABUSED; AND GIRLS WERE FREQUENTLY THE VICTIMS OF ABUSE. THERE WAS A TENDENCY FOR THE YOUNGEST CHILD TO BE MORE OFTEN ABUSED THAN OTHER SIBLINGS. PARENTS' PERCEPTIONS OF THE ABUSED CHILD WERE GENERALLY NEGATIVE, SUGGESTING THAT A CERTAIN INTERACTION BETWEEN THE PARENT AND A PARTICULAR CHILD TRIGGERS THE CHILD ABUSE. FURTHER RESEARCH IS RECOMMENDED TO VALIDATE THESE FINDINGS IN OTHER SAMPLES AND POPULATIONS. REFERENCES ARE INCLUDED. (JAP)