This study used structural equation modeling with longitudinal data from the Oregon Youth Study to test the hypothesis that the effect of socio-economic status (SES) on delinquency in early adolescence would be mediated entirely by parental management skills. SES was measured by parental education and occupation when the son was in the fourth grade, parental management skills during the sixth grade and delinquency during the seventh grade. The hypothesis was supported: The direct effect of SES on delinquency was not significant after controlling for parental management, which was modeled as a second-order factor consisting of parental monitoring and discipline. Implications are discussed for the theories of delinquency and for delinquency prevention and treatment.