A criticism of early findings related to sample selection subsequent to identification as juvenile offenders. Of 3,164 births enrolled in the Brown University cohort of the National Collaborative Perinatal Project, 13.6 percent appeared in Family Court charged with juvenile offenses prior to age 18. In this study, those delinquency cases were compared with controls at three age levels. At 8 months, based on the Bayley Scales of Mental and Motor Development, there were no significant differences in mental or motor development scores between delinquents and nonoffenders. At 4 years, Stanford-Binet scores were significantly lower among the delinquent sample. Similarly, at age 7, the delinquent sample scored lower on the Wechsler Intelligence Scale for Children (WISC) on both verbal and performance measures. Performance scores were consistently higher than verbal scores in both the delinquent and nonoffender groups. These findings support the view that intelligence scores at an early age are a factor in the vulnerability of children toward future delinquency, an observation which has relevance for treatment options of children and their families when school-related problems arise and later, if they enter the juvenile justice system. (Author abstract)