A retrospective review of 1316 neonates who had been evaluated in the Perinatal Laboratory of my institution during the years 1990 and 1991 was undertaken to study the relationship between birthweight for gestational age, maternal high-risk factors, and neonatal ponderal index with neonatal outcome. Patients were stratified as to gestational age at delivery and the presence of significant congenital abnormalities. Univariant analysis (chi square or t test, where appropriate) was done comparing neonatal outcome for neonates that were or were not small for gestational age (SGA), whose mother did or did not have high-risk factors, and for infants with normal or abnormal ponderal indices. Multivariant analysis was also done using logistic regression analysis. In infants without congenital anomalies delivering at term only the presence of maternal risk factors had prognostic significance, whereas both maternal risk factors and defining infants as SGA had prognostic value in infants delivering preterm, especially less than 34 weeks' gestation. Ponderal index had no significant prognostic value at any gestational age. The data supports the concept that intrauterine growth retardation as defined as an SGA, low birthweight for gestational age infant, may be an inaccurate classification for infants delivering at term, but does have prognostic significance in preterm infants, especially those less than 34 weeks' gestational age.