A sample of 1,617 12- to 15-year-old boys were classified according to the presence or absence of LD and interviewed individually concerning family background, school attitude, and self-reported delinquent behavior. Two groups of boys were included in the sample: 633 boys who had been adjudicated delinquent by juvenile courts, and 984 boys who had no records of previous adjuducation. The percentages of boys in the delinquent and nondelinquent groups who were classified as learning disabled were 18.9 and 36.5, respectively. In a hierarchical multiple regression analysis, a general self-reported delinquency (SRD) measure was regressed on LD classification controlling for the effects of potentially confounding background variables. The results indicated that the presence of LD had a small, but significant negative relationship with general SRD. However, a strong positive relationship was found between the presence of LD and the probability of adjudication when adjudication status was regressed on LD classification in a logistic regression analysis. These results suggest that learning disabled youths do not evidence more delinquent behavior than nonlearning disabled youths but are more likely to be found delinquent by juvenile courts. Notes, footnotes, tables, and approximately 35 references are provided. (Author abstract modified)