OBJECTIVE: The aim of the study was to compare clinical and ultrasonographic variables obtained before delivery as predictors of neonatal survival and morbidity in infants weighing ≤1000 g at birth. STUDY DESIGN: Maternal data available before the birth of singleton infants with birth weights ≤1000 g who were delivered at the 11 tertiary perinatal centers of the National Institute of Child Health and Human Development Network of Maternal-Fetal Medicine Research Units were studied. Births that followed extramural delivery, antepartum stillbirths, multiple gestations, induced abortions, infants with major malformations, and fetuses delivered at <20 weeks' gestation were excluded. Ultrasonographic variables, including estimated fetal weight, obstetrically estimated gestational age, femur length, and biparietal diameter, and clinical variables, such as maternal race, antenatal care, substance abuse, medical treatment, reason for delivery, fetal gender, and presentation, were studied with logistic regression as predictors of neonatal outcome, including intrapartum stillbirth, neonatal death, and survival to 120 days after birth or to discharge from the hospital with or without the presence of markers of major morbidity. RESULTS: Eight hundred eight infants met enrollment criteria; 63 were excluded because of incomplete data and 32 were excluded because of malformations, leaving 713 for analysis, 386 of whom had an ultrasonographic examination within 3 days of delivery that recorded femur length, biparietal diameter, and estimated fetal weight. Forty-two percent of births were the result of preterm labor, 22% were the result of preterm ruptured membranes, 12% were the result of preeclampsia or eclampsia, 9% were the result of fetal distress, 4% were the result of placenta previa or abruptio placentae, and 2% were the result of intrauterine growth restriction. Perinatal mortality before 24 weeks' gestation exceeded 81% (19% stillbirths and 62% neonatal deaths) but declined sharply thereafter. Most survivors born before 26 weeks' gestation had serious morbidity. Fetal femur length and estimated gestational age predicted survival better than did biparietal diameter or estimated fetal weight. Infants who survived with markers of serious long-term morbidity could not be distinguished from those who survived without morbidity markers before delivery by ultrasonography or clinical data. Threshold values for ultrasonographic measurements of biparietal diameter and femur length were developed to distinguish fetuses with no chance of survival. CONCLUSION: Ultrasonographic assessment of either fetal femur length or gestational age predicts neonatal mortality better than do other antenatal tests. No tests accurately predicted neonatal morbidity in infants weighing ≤1000 g at birth.