In order to test this theory, data were collected from questionnaires administered to 724 students in four high schools and three juvenile correctional institutions in the Southwest. Findings from this study indicate that when subjected to path analysis, Hirsch's contention is only supported when delinquent companions is included; that is to say the extent of explained variation in delinquency is greatly enhanced when delinquent companions is introduced as an additional antecedent variable in Hirsch's causal scheme. In addition, the findings are more consistent with a social learning or differential association theory than the original theory proposed by Hirschi in Causes of Delinquency (1969). (Author abstract)