A retrospective analysis of 239 postdate patients who delivered within seven days of their last nonstress test between 1 July 1980 and 30 June 1981 was done to determine whether or not there was a relationship between variable fetal heart rate (FHR) decelerations, a decline in the FHR of at least 15 beats per minute and lasting 15 seconds, on the nonstress test and pregnancy outcome. Of the 505 nonstress tests, the results were considered either reactive (424 to 84%) or nonreactive (81 to 16%). Variable FHR decelerations or bradycardias were encountered on 95 nonstress tests (18.8%) in 80 (33.5%) postdate patients. Fetal heart rate bradycardia was encountered in nine patients (3.8%). Based on the last nonstress test, pregnancy outcome for the nonreactive group was less favorable than for the reactive group. However, when the reactive group was separated as to whether FHR decelerations or bradycardias were present or absent on the last test, the reactive group with FHR decelerations had a significant increase in meconium passage, cesarean delivery for fetal distress, and Apgar scores less than 7 at one minute. This group also had a similarly poor outcome as the nonreactive group of postdates patients. These data suggest that the nonstress test, when reactive without evidence of FHR decelerations, is a reliable indicator of fetal well-being in the postdate pregnancy. However, a reactive nonstress test with evidence of FHR decelerations is associated with a significant increase in perinatal morbidity.