The role of nonstressed monitoring of the fetal heart rate (HR) in determining fetal well-being during the antepartum period was assessed in 125 high-risk patients. Observations on HR, variability, and HR response to fetal movement (FM) and uterine contractions (UC) over a 30 minute period were made with an external microphone and tocotransducer. A total of 625 tests were performed; the earliest gestation tested was 28 weeks, and the latest was 46 weeks. A reactive pattern (variability greater than 6 b.p.m. and accelerations with FM) appears to be a reliable indicator of fetal well-being. All the 51 fetuses exhibiting this pattern survived. This group also had the lowest incidence of neonatal complications. On the other hand, of the babies who failed to show variability greater than 6 b.p.m. or accelerations with FM (nonreactive pattern), 40% died in the perinatal period. Thirty-five patients showed features of both a reactive and nonreactive pattern (combined pattern). Poor outcome in this group was confined to those in whom the majority of the pattern was nonreactive. An undulating HR pattern with virtually absent variability (sinusoidal pattern) was found in 20 Rh-sensitized fetuses, 50% of whom died in the perinatal period. Bradycardia and tachycardia were not found to be reliable signs of fetal distress antepartum. Of the 12 fetuses who died during observation, six showed late decelerations with spontaneous UC but all showed diminished variability. The close correlation between nonstressed patterns and neonatal outcome demonstrated by this preliminary study warrants further use of this technique for fetal evaluation.