Fetal responses to maximal maternal exercise were studied during cycle ergometry and tethered swimming in 13 untrained subjects at 25 and 35 weeks' gestation. The fetal heart rate (FHR) and uterine and umbilical artery waveforms were measured before exercise, immediately after exercise, and at 5-minute intervals during 20 minutes of recovery. The mean maternal maximal heart rate was 179 +/- 12 beats per minute and did not differ between swimming and cycling exercise trials. Six episodes of transient bradycardia occurred after a total of 45 maximal exercise trials. The mean FHR decreased slightly immediately after exercise, then increased approximately ten beats above baseline levels at 10-20 minutes after exercise (P less than .02). The changes from baseline in FHR were greater after cycling than after swimming. Changes observed in the systolic-diastolic ratio (S/D) of the umbilical artery could be accounted for by the FHR variations. The S/D of the uterine artery was significantly higher after cycling than after swimming (P = .05). We conclude that maximal maternal exercise during pregnancy causes transient fetal bradycardia in approximately 15% of cases, in cycling more often than in swimming.