It has been reported that determinations of maternal serum unconjugated estriol (E3) and estetrol (E4) concentrations provide clinicians with more or less identical information on fetal status. If this is true, then theoretically the levels of E3 should be equally correlated with those of E4 in all conditions of pregnancy. To resolve this question, a study of the relationship between E3 and E4 was performed before labor in normal and complicated pregnancies. In normal pregnancy, they were highly correlated (r = .683, P less than .0001); in complicated cases, they were still correlated, but at a lower level (r = .522, P less than .003). To determine the effect of labor on this correlation, E3 and E4 levels were measured in normal subjects during labor as well as in the corresponding fetuses. The correlations between material E3 and E4, maternal and fetal E4, maternal and fetal E3, and fetal E3 and E4 were highly significant. A similar study with complicated pregnancies, however, indicated no such correlation except between fetal E3 and E4. In addition, day-to-day variability of serial measurements of E3 and E4 on an individual basis was determined in normal and diabetic subjects. The variability was qualitatively determined graphically and quantitatively determined algorithmically. The results of calculated individual variabilities indicated that the variability of E4 was less than that of E3 in most cases. It is therefore concluded that complications in pregnancy and the onset of labor have some effect on E3-E4 correlations, and that measurement of E4 has an additional advantage due to less variability.