OBJECTIVE A prospective study comparing three management schemes for patients at term with premature rupture of membranes was performed. METHODS One hundred forty patients were randomized to one of three study groups: prostaglandin E2, placebo, or oxytocin. Patients randomized to prostaglandin E2 and placebo received vaginal suppositories containing 3 mg prostaglandin E2 or glycerin only, respectively; suppositories were administered in a double-blind fashion, on one or two occasions, 6 hours apart. Oxytocin was given only if labor was not established after 12 hours or to augment inadequate labor. In patients randomized to oxytocin labor was induced with intravenous oxytocin. The time interval to delivery, delivery outcome, and complications were analyzed. RESULTS Patients receiving prostaglandin E2 were more likely to be in labor after one suppository and to be delivered without the addition of oxytocin when compared with placebo. The time interval to delivery was shorter with prostaglandin E2 and oxytocin induction versus placebo ("expectant management"). The incidence of maternal infection was lowest in patients with labor induced by prostaglandin E2. Although the overall cesarean section rate was low, there was a trend toward a lower rate with prostaglandin E2 induction. No adverse effects were observed with prostaglandin E2. CONCLUSIONS Prostaglandin E2 can be used successfully to induce labor after premature rupture of membranes at term with greater ease of administration when compared with oxytocin.