THE CAUSAL MODEL USED INCLUDES MOST OF THE MAJOR VARIABLES THAT HAVE BEEN PROPOSED AS CAUSES OF DELINQUENT BEHAVIOR--SOCIAL CLASS, PARENTAL LOVE, SCHOOL PERFORMANCE, ATTACHMENT TO PARENTS, OCCUPATIONAL EXPECTATIONS, NEED FOR PEER APPROVAL, DELINQUENT ASSOCIATES, DELINQUENT VALUES (ABOUT THE ACCEPTABILITY OF CERTAIN ACTIONS), AND PERCEIVED RISK OF APPREHENSION. THE RELATIVE IMPORTANCE OF THE EXPLANATORY VARIABLES IS COMPARED FOR SEPARATE PATH ANALYSES OF THEFT, VANDALISM, AND ASSAULT, AND FOR THE COMBINED 'TOTAL' DELINQUENT ACTIVITY AS SELF-REPORTED BY SUBJECTS. RESULTS INDICATE THAT WHILE GENERAL SIMILARITIES IN THE APPARENT CAUSES OF THESE KINDS OF DELINQUENT ACTIVITIES EXIST, RESEARCH WHICH ATTEMPTS TO ANALYZE DELINQUENCY AS A COMBINED INDEX OF A VARIETY OF BEHAVIOR PATTERNS MAY MASK SOME IMPORTANT DIFFERENCES IN THE PROCESSES WHICH GENERATE SPECIFIC KINDS OF DELINQUENT ACTS. THE DATA SUGGEST, FOR EXAMPLE, THAT DELINQUENT ASSOCIATES AND SCHOOL PERFORMANCE ARE MORE SALIENT AMONG THE FORCES LEADING TO THEFT, WHILE PERSONAL VALUES ARE MORE IMPORTANT IN PREDICTING ASSAULTIVE ACTIVITIES. METHODOLOGY, TABULAR DATA, AND REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--DAG)