A 50 percent sample (n=9,335) of the 7th grade students in the Houston Independent School District (Texas) in 1971 responded to a self-administered questionnaire during the school day in the spring of that year (Tl). A virtually identical questionnaire was administered to wiling students who remained in school during the 8th and 9th grades during the spring of 1972 (T2) and 1973 (T3) respectively. The questionnaires measured aspects of self-rejecting attitudes and delinquent behavior. The elements of the model developed were the central concept of self-derogation, specified antecedents of self-derogation, and specified sequels of self-derogation. Four of the variables were scores at T1, and seven of the variables were expressed as residualized gains between T1 and T2. The outcome variables were self-reports at T3 with regard to performance of each of 28 deviant acts between T2 and T3. Bivariate analyses of the relationship between self-derogation at T1 and subsequent deviant responses supported the expectation that self-rejecting attitudes are a general antecedent of multiple modes of deviant responses. Regarding the putative antecedents of self-derogation as having effects upon the adoption of deviant responses, such deviant responses are generally anticipated by high felt rejection by the school and by the family, measured at T1. Viable peer relationships provide the occasion or resources for adopting several deviant responses. Increased awareness of deviant alternatives in the environment, as expected, increased the likelihood of adopting each of the deviant patterns. Tabular data and 30 references are provided.