The study shows that the main premise of GST--that increased sources of strain can lead to increased delinquent behavior--is supported by this study. Increased victimization, as a source of strain, was found to predict increased amounts of violent offending, illicit drug use, and nonviolent (nondrug) offending. The study's support of GST, however, is conditional when the analysis takes into account the results of the interaction of variables that condition strain. In terms of violent offending, increased victimization was an important predictor for both racial groups in the panel (Whites and African-Americans); however, victimization was found to increase violent offending slightly more over time for African-Americans than Whites. Regarding illicit drug use, victimization in adolescence had a greater effect on African-Americans than Whites. Victimization increased nonviolent, nondrug offending more in Whites than African-Americans. Regarding gender, victimization was found to be predictive of violent offending for boys but not for girls. Victimization was apparently more important in predicting illicit drug use for boys than for girls. Victimization increased nonviolent, nondrug offending more for girls than for boys. This study examined how these findings differed from previous similar studies and what might explain the differences. Data used in this study came from the National Youth Survey, a current, ongoing longitudinal study that began in 1976. From the original 1,725 respondents, only those who were White or African-American were included in the current study. 8 tables, 112-item bibliography, appended missing data analyses, and a subject index