ACCORDING TO HIRSCHI'S THEORY OF THE CAUSES OF DELINQUENT BEHAVIOR, THIS BOND CONSISTS OF FOUR ELEMENTS: ATTACHMENT, COMMITMENT TO CONVENTIONAL LINES OF ACTION, INVOLVEMENT IN CONVENTIONAL ACTIVITIES, AND BELIEF IN THE EXISTING RULES. HIRSCHI TESTED HIS THEORY UTILIZING A SAMPLE OF WHITE MALE ADOLESCENTS AND FOUND SUPPORT FOR ALL EXCEPT ONE OF THE HYPOTHESIZED ELEMENTS OF THE BOND: INVOLVEMENT IN CONVENTIONAL ACTIVITIES. HE ALSO UNDERESTIMATED THE INFLUENCE OF DELINQUENT PEERS. THIS DISSERTATION REPORTS ON A REPLICATION OF HIRSCHI'S ANALYSIS USING A LONGITUDINAL RESEARCH DESIGN. THE SAMPLE CONSISTED OF 2617 MALES AND FEMALES OF DIVERSE CLASS AND ETHNIC BACKGROUND. MEASURES OF THE INDEPENDENT VARIABLES WERE OBTAINED WHEN THE COHORT WAS IN THE NINTH GRADE (TIME ONE). SUBSEQUENT DELINQUENT BEHAVIOR, THE DEPENDENT VARIABLE, WAS MEASURED BY A SELF-REPORTED DELINQUENCY CHECKLIST ADMINISTERED IN THE LATTER PART OF THE COHORT'S SENIOR YEAR (TIME TWO). THE FINDINGS OF AN INTENSIVE REPLICATION WERE CONSISTENT WITH THE RESULTS OF HIRSCHI'S ANALYSIS. IN AN EXTENSIVE REPLICATION, A LARGE POOL OF ITEMS THOUGHT TO MEASURE SOCIAL CONTROL VARIABLES WERE FACTOR ANALYZED TO DERIVE SCALES. A REGRESSION OF THESE SCALES ON SUBSEQUENT DELINQUENT BEHAVIOR PROVIDED LITTLE SUPPORT FOR SOCIAL CONTROL THEORY. IN FACT, THE CONTROL THEORY VARIABLES WERE NOT AS EFFICIENT AS PREDICTORS OF SUBSEQUENT DELINQUENT BEHAVIOR AS WAS KNOWLEDGE OF PRIOR INVOLVEMENT IN DELINQUENCY. A FINAL ANALYSIS OF A RESTRICTED PORTION OF THE SAMPLE (STUDENTS WHO HAD NOT REPORTED SERIOUS DELINQUENT ACTS AT TIME ONE) PROVIDED MODERATE SUPPORT FOR THE IDEA THAT ATTACHMENT TO PARENTS AND TO THE SCHOOL HAVE A PREVENTIVE INFLUENCE ON DELINQUENCY. IT WAS CONCLUDED THAT THE PREDICTIVE ABILITY OF SOCIAL CONTROL THEORY WAS SIMILAR ACROSS CLASS AND ETHNIC GROUPS, BUT NOT AS EFFICIENT IN PREDICTING FEMALE DELINQUENCY AS IT IS MALE DELINQUENCY. A 19 PAGE BIBLIOGRAPHY IS INCLUDED. (AUTHOR ABSTRACT MODIFIED)