Newborn status, ratings of temperament, and heretofore neglected experiential (parenting) antecedents of hyperactivity were evaluated in a prospective, longitudinal investigation. The experiential variables were derived from a developmental theory of arousal modulation in the early years. Patterns of maternal intrusive care, seductive behavior, and overstimulation were assessed at 6 months, 2 years, and 3 1/2 years, respectively. 2 of these 3 variables significantly predicted hyperactivity in kindergarten at age 5 or 6, as determined by clinically validated ratings made by teachers. Of the 38 early child variables, which included neonatal behavioral assessments and ratings of activity level by parents and observers on many occasions during the first 2 1/2 years, only the motor maturity factor from 2 composited Brazelton exams yielded a significant relation. In addition, a measure of distractibility at 42 months was predictive of hyperactivity. Although other child measures may prove to be more powerful than those we tested, it will remain the case that prospective data are a critical arbiter of etiological factors in hyperactivity. Such data have important implications for both diagnosis and treatment. [ABSTRACT FROM AUTHOR]