Over the past decade, 24-hour urinary estriol determination has become an integral part of the evaluation and monitoring of high-risk obstetric patients for fetal well-being. This study was undertaken to assess the utility of estriol determinations obtained by simpler collection methods: serum estriol and first morning urine specimens. Serum estriol proved to be an unreliable predictor of 24-hour urine values and of estriol fluctuations. Although the correlation between first morning and 24-hour primary estriol:creatinine ratios was statistically significant (P less than .0001), any single first morning specimen value chosen at random was a poor predictor of the corresponding 24-hour specimen value. However, in the monitoring of high-risk obstetric patients, it is the significant changes with respect to time rather than the absolute estriol levels that are of interest. As the plots of first morning and 24-hour specimen values with respect to time in 8 subjects exhibit time trends of equivalent diagnostic utility, it it concluded that first morning values may be used in place of 24-hour values to monitor high-risk obstetric patients, making the process simple and quicker.