PREVIOUS STUDIES HAVE SHOWN THAT CHILD ABUSE IS RELATED TO YOUNG AGE OF PARENT AND TO ATTITUDES ABOUT CHILDREARING HELD BY THE PARENT. APPROPRIATE INTERVENTION, THEREFORE, MUST BEGIN DURING EARLY ADOLESCENCE BECAUSE MANY ABUSING PARENTS HAVE THEIR FIRST CHILD BEFORE THEY ARE OUT OF THEIR TEENS. A QUESTIONNAIRE WAS DEVELOPED TO IDENTIFY 'HIGH RISK' ADOLESCENTS, THOSE WHOSE ATTITUDES TOWARDS PARENTING INDICATE A HIGH POTENTIAL FOR CHILD ABUSE. IT WAS BUILT AROUND FOUR AREAS PREVIOUS RESEARCHERS HAVE FOUND TO BE SIGNIFICANT: (1) INAPPROPRIATE PARENTAL EXPECTATIONS FOR THE CHILD, (2) INABILITY OF THE PARENT TO BE EMPATHICALLY AWARE OF THE CHILD'S NEEDS, (3) STRONG PARENTAL BELIEF IN THE VALUE OF PUNISHMENT, AND (4) ROLE REVERSAL. THREE PROTOTYPES WERE DEVELOPED AND FIELD TESTED. EACH PROTOTYPE QUESTIONNAIRE, THE STATISTICAL DATA RESULTING FROM THE VALIDATIONS, AND THE FINAL QUESTIONNAIRE ARE GIVEN. THE FINAL VERSION IS BASED ON 32 ITEMS WHICH HAD THE HIGHEST POSITIVE FACTOR LOADINGS (EQUAL OR GREATER THAN 0.20). THE RANGE OF ITEM-CONSTRUCT CORRELATIONS FOR THESE ITEMS WAS 0.53 TO 0.75. THE ITEMS SHOWED HIGH INTERNAL CONSISTENCY AND GOOD STABILITY WHEN THE INSTRUMENT WAS ADMINISTERED 1 WEEK LATER. ABUSED ADOLESCENTS SCORED SIGNIFICANTLY LOWER ON PARENTING ATTITUDES (P LESS THAN 0.001) AND MALES LOWER THAN FEMALES (P LESS THAN 0.001). CAUTIONS IN USE OF THE INSTRUMENT ARE GIVEN. REFERENCES ARE APPENDED. (GLR)