This is a prospective study of differential morbidity among subgroups of intrauterine growth retardation. Cases of intrauterine growth retardation (N = 3450) (greater than or equal to 37 weeks, less than 10th percentile birth weight for gestational age) were classified by their ponderal index (weight/length3) in four subgroups using the 10th, 25th, and 90th percentiles of the Lubchenco's ponderal index-gestational age distribution. There were 432 cases (12.5%) with low ponderal index or disproportionate intrauterine growth retardation, 936 (27.1%) with intermediate ponderal index, 2030 (58.8%) with adequate ponderal index or proportionate intrauterine growth retardation, and 52 (1.5%) with high ponderal index. The low ponderal index group or disproportionate intrauterine growth retardation group had a statistically significant higher risk (between 1.6 and 12.5 times) for low 1- and 5-minute Apgar scores, aspiration syndrome, hypoglycemia, and perinatal asphyxia than the adequate ponderal index group. The low ponderal index group also had an increased risk (relative risk = 2.0 [95% confidence interval, 1.0 to 3.8]) for hospital stay of more than 1 week. These differences persist after a stratified analysis by birth weight and in a multiple logistic regression analysis. Similarly, higher neonatal morbidity is observed among infants with normal birth weights but with low ponderal index. These data provide further evidence of the heterogeneity of the intrauterine growth retardation syndrome and of the independent effect of body disproportion on neonatal morbidity, even among infants with normal birth weights. Because there are significant clinical implications attributed to the low ponderal index group, this subgroup should be identified as early as possible.