The hypothesis that sustained uterine contractile activity is the direct cause of fetal death after progesterone withdrawal in late pregnancy in rats was investigated. Pregnant rats were subjected to progesterone withdrawal on day 15 of pregnancy by injecting 2 mg mifepristone (RU 486) kg-1 or by ovariectomy with oestradiol replacement (200 ng day-1). Uterine contractile activity (force and frequency) at 4 h, but not at 2 h, in rats injected with mifepristone was significantly higher than in rats injected with vehicle. The contractile activity in mifepristone-treated rats remained higher than in control rats, at 12, 24 and 48 h. Fetal viability 36 h after mifepristone injection, when uterine contractions had lasted for 32 h, was not significantly different from fetal viability in rats injected with vehicle, but at 42 h after mifepristone injection, fetal viability was significantly reduced. In ovariectomized rats, uterine contractile activity at 12, 24, 36 and 48 h, but not at 8 h, was significantly greater than in ovariectomized rats with progesterone replacement (4 mg day-1). Fetal viability at 42 h after the operation, when uterine contractions had lasted for 30 h, was not significantly reduced, but it was significantly reduced at 48 h. When ovariectomized rats had been left to develop uterine contractions for a period before progesterone was injected, deprivation of progesterone and prolonged uterine contractions for about 30 h did not reduce fetal viability or fetal growth determined on day 18, but it did so 3 days later, on day 21. Administration of 5 mg isoxuprine kg-1 twice a day, which suppressed uterine contractions, improved fetal viability in ovariectomized rats at the earlier stage, but not at the later stage. Nevertheless, isoxuprine did improve fetal growth at the later stage in these ovariectomized rats. It is concluded that increased uterine contractile activity sustained for 32 h or less does not reduce fetal viability, but longer periods of contraction may be the cause of fetal death.