EEG'S WERE RECORDED FROM 35 ENGLISH SUBJECTS WHO EITHER CONFESSED TO CHILD ABUSE OR IN WHOM THE INDEX OF SUSPICION WAS HIGH ENOUGH TO MAKE IT VIRTUALLY CERTAIN THAT THEY HAD DONE SO. THE SPOUSES OF 16 OF THESE SUBJECTS WERE ALSO SUBJECTED TO EEG EXAMINATION. THE RESULTS SHOWED THAT ALMOST 25 PERCENT OF THE PARENTS WHO BATTERED THEIR CHILDREN HAD DEMONSTRABLE ABNORMAL EEG'S. THOSE IN THIS SUBGROUP WERE GENERALLY FOUND TO BE PSYCHOPATHIC AN OF LOW INTELLIGENCE, ALTHOUGH NO LOWER ON AN AVERAGE THAN THOSE IN WHOM NO EEG ABNORMALITY COULD BE SHOWN. WITHOUT REFERENCE TO THE ACT OF BATTERING, THIS SUBGROUP COULD ALSO BE DEFINED AS HAVING A PERSONALITY DISORDER ACCORDING TO THE AMERICAN PSYCHIATRIC ASSOCIATION CLASSIFICATION (1952). THEY WERE ALSO FOUND TO BE PERSISTENT BATTERERS, HAVING BATTERED ONE CHILD MORE THAN ONCE, AND SOMETIMES MORE THAN ONE OF THEIR CHILDREN. THE INABILITY TO TREAT CHILD ABUSERS AS A HOMOGENOUS GROUP IS EXPLAINED. THE PAPER SUPPORTS THE CONCLUSION THAT IT IS NOT SAFE TO GENERALIZE ABOUT WHY CHILD BATTERERS PURSUE THIS BEHAVIOR PATTERN. THE SUBGROUP WITH ABNORMAL EEG'S SUGGESTS AN ORGANIC THEORY THAT NEEDS CLOSE ATTENTION. A RELATED DISCUSSION INCLUDES A CLASSIFICATION OF FOUR TYPES OF ABUSE CASES (TRUE INFANTICIDE, WASTED AND NEGLECTED, DELIBERATE CRUELTY, AND BATTERING) AND REPORTS RESULTS FROM THE TESTING OF A CONTROL GROUP OF CHILDREN ADMITTED TO THE EMERGENCY ROOM OF AN ENGLISH HOSPITAL. (RBS)