To assess the value of prophylaxis with penicillin in women with premature rupture of membranes (PROM) and the long-term outcome of children born after prolonged PROM, we studied 221 women with this condition. Penicillin (5 mu twice, 6 hours apart) was given intravenously to 50 women and placebo to 51 women, whereas 76 comparable patients were treated without penicillin or placebo. The time interval between PROM and delivery ranged from 14 hours to 56 days. Chorioamnionitis occurred more frequently (p less than 0.05) among patients with placebo (14%) than among those treated with penicillin (2%). One puerperal endometritis appeared in the placebo group compared with none in the penicillin group. One newborn (1.7%) born to a mother with placebo prophylaxis developed septicemia, compared with none in the penicillin group. The outcomes of pregnancies complicated with PROM treated without penicillin or placebo were comparable with those in the placebo group. In addition, we compared somatic and psychomotor development of 159 children born to mothers with prolonged PROM (more than 12 hours; mean, 5.6 days; range, 14 to 1344 hours) with those of 43 children born at similar gestational age within 5 hours after PROM. No pulmonary sequelae could be linked to the long time period between PROM and delivery, but infants born soon after PROM more often (p less than 0.05) had cerebral palsy (8 of 43, 18.6%) than did infants born after prolonged PROM (7 of 159; 4.4%). We conclude that, in cases with PROM, penicillin prophylaxis decreases maternal and neonatal infectious morbidity and that the long interval between PROM and delivery does not impair the long-term outcome for these children.