To determine whether the addition of broad-spectrum antimicrobial therapy to traditional expectant management improves pregnancy outcome in patients with premature rupture of membranes (PROM) remote from term. Methods: Patients with preterm PROM before 34 weeks’ gestation who were not in labor and had no signs of infection or fetal distress were randomized to one of two study groups: 1) expectant management alone and 2) expectant management plus antimicrobial therapy. Women in the latter group received intravenous ampidllin, gentamicin, and clindamycin for 24 hours, followed by amoxicillin plus clavulanic acid orally for 7 days. Other than antibiotic use, management of the two groups was identical. Results: Significantly more women (P <.01) treated with antibiotics (20 of 48, 42%) remained undelivered 7 days after admission compared with those managed expectantly without antibiotics (seven of 46, 15%). In addition, more neonates in the group managed with antibiotics were admitted to the routine nursery (nine of 48 versus two of 45; P =.03). However, there was no difference between the groups in the frequency of serious maternal or neonatal morbidity. Conclusions: The addition of broad-spectrum antimicrobial therapy to traditional expectant management of pregnancy complicated by preterm PROM may increase the number of gestations undelivered 7 days after admission. It may also decrease the proportion of infants, admitted to special care nurseries. Whether these effects result in significant short-or long-term maternal or neonatal benefit remains to be determined. © 1992 by The American College of Obstetricians and Gynecologists.