Two popular criminological theories, social control theory and general strain theory (GST), both make predictions on how attachment to parents and teachers will affect delinquency in youth. Their predictions differ somewhat; social control theory holds that positively attached youth will be less delinquent than neutral and negatively attached youth and that there will be little difference in the delinquency of neutral and negatively attached youth. On the other hand, GST contends that negatively attached youth will be more delinquent than neutral and positively attached youth and that there will be little difference in the delinquency of neutral and positively attached youth. The current study tested these competing theories using data from the national “High School Youth, Weapons, and Violence” survey conducted of 734 male high school sophomores and juniors from 134 randomly selected schools in 1996. Results of ordinary least squares regression analysis and polynomial regression models indicate that both parental and teacher attachment have a nonlinear effect on delinquency and the effect supports the predictions made by GST. In sum, negatively attached youth were significantly more delinquent than were neutral and positively attached youth. Neutral and positively attached youth differed little in their rates of delinquency. These findings suggest that the dominant explanation for the effect of attachment on delinquency, namely social control theory, may be incorrect, although the results here should be viewed as tentative given the small sample size and limited number of negatively attached boys in the sample. Future research should continue to investigate the nonlinear effects of relevant variables on criminality. Figures, tables, references, appendixes