The cesarean delivery rate has quadrupled during the past two decades, resulting in considerable attention focused on alternatives to cesarean birth. One option, vaginal birth after one previous cesarean, has come to be recognized as an acceptable alternative to routine elective repeat cesarean delivery. The purpose of this report was to evaluate whether women with two previous cesareans can safely undergo a trial of labor. Between July 1, 1982 and June 30, 1986, data were collected prospectively on all women with previous cesareans. Those with a known classical incision or a medical or obstetric contraindication to a trial of labor were excluded from an attempted vaginal delivery. During this period, 67,784 patients were delivered, of whom 6250 (9.2%) had had a previous cesarean. Of the 6250 previous-cesarean patients, 1088 (17.4%) had had two previous cesareans; of these, 501 (46%) underwent a trial of labor and 346 (69%) delivered vaginally. Whereas the overall rate of uterine dehiscence was 3%, the rate in those women who attempted a vaginal delivery was 1.8%, versus 4.6% in those who did not. Overall, oxytocin was used in 284 (57%) and was associated with a dehiscence rate of 2.1%, versus 1.4% in the no-oxytocin group. Successful vaginal delivery was related significantly to the use of oxytocin and to a previous vaginal delivery. Trial of labor in patients with two previous cesareans appears to be a reasonable consideration.