A retrospective study of the influence of spontaneous preterm labor on perinatal mortality rate over a period of 15 years was conducted at the National Maternity Hospital, Dublin. As the perinatal mortality rate from all causes declined from 48 to 16 per 1,000 births, the rate attributed to spontaneous preterm labor declined in the same proportion, from 10 to three per 1,000 births. Review of the circumstances of death attributed to spontaneous preterm labor in each case indicates that improvement in this area was due mainly to a natural reduction in incidence and, to a lesser degree, to better care of the neonate. Pharmacologic agents recommended for the purpose of averting spontaneous preterm labor made no contribution, because none was used. Natural improvement in the evolution of a disease may lead to exaggerated claims for the benefits of treatment; experience with perinatal death attributed to spontaneous preterm labor in this large unit suggests that current enthusiasm for tocolytic agents may well be misplaced.