Predicting developmental outcome from infancy is problematic, with low correlations usually reported between developmental scores in infancy and later IQ or achievement. Further, little is known about the stability of developmental measures of very premature infants over time. Rosenbaum et al.(Devel. Med. Child Neurol., 1995) suggested that adaptive behavior, or growth toward personal self-sufficiency, may be more descriptive of outcome than other conventional measures. The purpose of the present study was to investigate the stability of mental and adaptive development in a cohort of very low birth weight (VLBW) children. The subjects were 50 VLBWchildren born between 1986 and 1988. Mean gestation was 27.5 weeks; mean birth weight was 940 gm. As part of a longitudinal follow-up study, each child was evaluated at approximately 1, 4,and 8 years of age. At 1 year (adjusted for gestation age), children were evaluated using the Bayley Scales of Infant Development and the Vineland Adaptive Behavior Scales. At 4 years, the Wechsler Preschool and Primary Scale of Intelligence - Revised (WPPSI-R) and the Vineland were used. At 8 years, each child was administered the Wechsler Intelligence Scale for Children - Third Ed. (WISC-III) and the Vineland. Repeated measures ANOVAs found significant effects for subject age when comparing scores of cognitive(Bayley, WPPSI-R, WISC-III; F= 22.0, p= <.001) and adaptive development(Vineland; F= 11.78, p= <.001). Post-hoc analyses revealed cognitive scores at each age to differ significantly from one another. Vineland scores at age 1 were higher than both 4 and 8 years, which did not differ. Correlations between cognitive measures at each age were moderate to very high (r 1,4=.45; 1,8=.55; 4,8=.83, p= <.001), as were correlations between adaptive measures at each age (r 1,4=.48; 1,8=.60; 4,8=.68,p= <.001). While parent report of children's adaptive development was consistent across time, scores reflecting cognitive development, while somewhat predictive in terms of correlations, differed significantly in mean values across ages. One implication of these findings is that clinicians need to put increased emphasis on reporting of adaptive development when evaluating at-risk children.