The relationship between genital tract infection and preterm delivery has been established on the basis of biochemical, microbiological, and clinical evidence. In theory, pathogenic bacteria may ascend from the lower reproductive tract into the uterus, and the resulting inflammation leads to preterm labor, rupture of the membranes, and birth. A growing body of evidence suggests that preterm labor and/rupture of the membranes are triggered by micro-organisms in the genital tract and by the host response to these organisms, ie, elaboration of cytokines and proteolytic enzymes. Epidemiologic and in vitro studies do not prove a cause-and-effect relationship between infection and preterm birth. However, the preponderance of evidence indicates that treatment of asymptomatic bacteriuria and symptomatic lower genital tract infections such as bacterial vaginosis (BV), trichomoniasis, gonorrhea, and chlamydia will lower the risk of preterm delivery. Based on current evidence, pregnant women who note an abnormal vaginal discharge should be tested for BV, trichomonas, gonorrhea, and chlamydia. Those who test positive should be treated appropriately. A 3- to 7-day course of antibiotic treatment for asymptomatic bacteriuria during pregnancy is clinically indicated to reduce the risk of pyelonephritis and preterm delivery. Routine screening for chlamydia and gonorrhea should be performed for women at high risk of acquiring sexually transmitted diseases. The practice of routine screening for BV in asymptomatic women who are at low risk for preterm delivery cannot be supported based on evidence from the literature. Routine screening for asymptomatic bacteriuria during pregnancy is cost-effective, particularly in high-prevalence populations. The results of antibiotic trials for the treatment of preterm labor have been inconsistent. In the absence of reasonable evidence that antimicrobial therapy leads to significant prolongation of pregnancy in the setting of preterm labor, antibiotics should be used only for protecting the neonate from group B streptococci sepsis. They should not be used for the purpose of prolonging pregnancy. Multiple investigations have shown that, in patients with preterm premature rupture of the membranes, prophylactic antibiotics are of value in prolonging the latent period between rupture of the membranes and onset of labor and in reducing the incidence of maternal and neonatal infection. The most extensively tested effective antibiotic regimen for prophylaxis involves erythromycin alone or in combination with ampicilln. Controversy still exists regarding the appropriate length and route of antibiotic prophylaxis.