The nonstress test (NST) and the contraction stress test (CST) have had wide application to term pregnancies, but little has been reported of use in the management of preterm fetuses. Seventy-two of 438 high-risk gravidas, receiving both tests between 25 and 34 weeks' gestation, delivered singleton infants before completion of their 34th week. On the basis of the last NST and CST preceding delivery, the 72 fetuses could be divided into reactive-negative and nonreactive-positive groups. Nonreactive-positive fetuses experienced greater perinatal mortality and significantly higher rates of intrapartum fetal distress, neonatal depression, respiratory distress syndrome (RDS), intrauterine growth retardation, and cesarean section. Pulmonary complications in nonreactive-positive fetuses, regardless of the lecithin: sphingomyelin ratio, were significantly increased if intrapartum fetal distress had preceded delivery. Most neonatal deaths stemmed from RDS-related complications, 86% preceding the 31st week of gestation; however, nearly half of the perinatal deaths before the 30th week followed reactive NSTs. The data suggest that for clinical management of preterm pregnancy, fetal heart rate testing should be initiated after the 29th week; earlier, test significance is less clear and infant survival less likely. Maternal estriol determinations aid little in the management of nonreactive-positive fetuses. Rather, the care of these selected pregnancies should be temporized through the 30th week; when delivery is then elected, intrapartum fetal distress should be avoided through liberal use of cesarean section in the nonreactive-positive group.