To determine the effect of early postpartum discharge (less than 48 hours after vaginal birth or 96 hours after cesarean delivery) on maternal and neonatal complications, maternal concerns, patient satisfaction, and cost savings. We performed a MEDLINE search of English-language journals for pertinent articles published from 1966 through January 1997. We also reviewed reference lists in all the articles retrieved in the search as well as those of major obstetric texts. We included all studies describing early postpartum discharge. Studies included five randomized controlled trials, ten cohort studies, one casecontrol study, and 12 case-series reports. We classified the data using the rating system of the U.S. Preventive Services Task Force. We calculated relative risks and 95% confidence intervals for maternal and neonatal readmission and outpatient treatment after early postpartum discharge. Most studies did not show an increase in maternal or neonatal morbidity after early discharge. The five randomized controlled studies did not meet criteria for properly designed trials. Most evidence consists of cohort studies and case-series (class II-2 and III evidence) of highly selected patients with extensive supplemental antepartum and postpartum care and education. The current data do not support or condemn widespread use of early postpartum discharge in the general population (class C recommendation). Early postpartum discharge appears safe for carefully selected, consenting patients. Whether these data can be extrapolated to the general population of pregnant women remains unknown.