A diagnosis of intrauterine growth retardation (IUGR) was made in 150 pregnancies on the basis that the estimated fetal body weight was less than -1.5 SD of the intrauterine growth curve for the Japanese population in two consecutive measurements. The estimation of fetal body weight was made from the biparietal diameter and abdominal circumference using the formula of Shepard et al. Maternal plasma estriol, urinary estriol and plasma hPL were determined, and routine instructions to take daytime bedrest and a high protein diet were made. 75 optionally selected cases were given oral allylestrenol 30 mg/day until the time of delivery (the medicated group), and the other 75 cases (control group) were given no medication. Ultrasonic and biochemical measurements were repeated every two weeks thereafter. The estimated fetal weight calculated from Shepard's formula was of sufficient accuracy to make it possible to predict the fetal weight before delivery. In the control group, the estimated fetal weight averaged 1,281 grams at the initial measurements at an average of 32.7 weeks' gestation when the diagnosis of IUGR was made, and 2,498 grams at the final measurements at an average of 38.7 weeks gestation immediately before delivery. In the medicated group, the estimated fetal weight averaged 1,242 grams at the initial measurements at 32.8 weeks' gestation, and 2,826 grams at the final measurements at 39.0 weeks' gestation. Statistically significant increases in the fetal body weight were noted in the medicated group. Maternal plasma estriol, urinary estriol and plasma hPL concentrations rose significantly in the medicated group when compared with the control group.(ABSTRACT TRUNCATED AT 250 WORDS)