a postterm pregnancy that was assessed weekly with nonstress testing. Small and colleagues (1987) reported that twice-weekly onstress testing in a corrected perinatal mortality rate of 4.3 per 1,000. and associates (1983) that natal mortality in postterm pregnancies was 2.1 per 1,000 using urinary estriol/creatinine ratios, 6.5 per 1,000 using the conaction stress test, and 24 per 1,000 using the nonstress test. Others have reported similar experiences, and outcomes tend to be worse when the interval between tests was as long as 1 week is especially true with the nonstress test (Barrett and associates, 1982), but it also appears to be true when fetal biophysical profile testing is used It now is appreciated that fetal heart rate decelerations ob served during nonstress testing are predictive of increased fetal and neonatal morbidity and mortality in postterm pregnancy (Benedetti and Easterling, 1988). Small and associates (1987), as well as many others, recommend induction in these instances, regardless of whether there are accompanying normal fetal heart accelerations. Divon and co-workers (1988) described similar findings using the biophysical profile described below. It seems likely that these decelerations are the result of diminished am nionic fluid that predisposes to cord compression