The number of children at risk for delays in motor and social development (MSD) associated with preterm delivery and low birthweight is increasing, but such children are generally not seen as being in need of evaluation. The objective of these analyses was to determine whether there are independent effects of birthweight and gestational age on MSD and the magnitude of effects. Subjects were a representative sample of 4621 US-born singleton children, aged 2-47 months, examined in the third National Health and Nutrition Examination Survey (1988-94). MSD was assessed using an age-appropriate scale. Birthweight and gestational age were taken from birth certificates. Mexican-American and 'other' race/ethnicity (other than non-Hispanic white, non-Hispanic black or Mexican-American), low parental education level, older maternal age, higher birth order, low birthweight (LBW, <2500 g) and preterm delivery (<37 weeks) were all found to be associated with significant (P < 0.01) delays in MSD. Three per cent of the infants and children were preterm LBW and 2.2% term LBW (<2500 g, 37-44 weeks). Adjusting for socio-demographic factors, preterm LBW children had lower MSD scores (-1.5 +/- 0.3 points, P < 0.0001) through early childhood, as did term LBW children (0.8 +/- 0.4 points, P < 0.03). For females, LBW was the most important perinatal predictor of a lowered score (-0.9 +/- 0.3 points compared with normal birthweight, P < 0.04). For males, scores were additionally decreased by -0.1 +/- 0.03 points/week (P=0.001) of early delivery. LBW children had less muscle mass, but adjusting for muscularity did not diminish the effects of birth size on MSD. LBW status and preterm delivery are associated independently with small, but measurable, delays in MSD through early childhood and should be considered along with other known risk factors for development delays in determining the need for developmental evaluation.