A practical management protocol to handle the large volume of post-term gestations at the Los Angeles County/University of Southern California Medical Center was introduced and evaluated in 880 patients seen between March 1, 1979, and February 29, 1980. The protocol divided patients into 2 groups based on substantiated "good" or unsubstantiated "poor" obstetric dates. After clinical examinations, the primary screening test used was twice weekly plasma unconjugated estriol (E3) determinations. Antepartum fetal heart rate testing, in the form of nonstress tests, was used initially in the good obstetric dates group, and then done twice weekly only if the E3 value was abnormal (less than 18 ng/ml with good obstetric dates and less than 12 ng/ml with poor obstetric dates). Patients with good obstetric dates were delivered electively at more than 42 weeks' gestation if the cervix was favorable (Bishop score 9 or greater); otherwise, intervention occurred only with abnormal tests and a positive or suspicious contraction stress test, or with other medical indications. Only 8 perinatal losses (3 neonatal deaths and 5 stillbirths) occurred in 880 patients. Each of these patients received a follow-up evaluation: 3 had severe congenital anomalies, and 5 deaths occurred in patients who did not comply with the protocol. The cesarean section rate was 15.8%.