The relationship between asymptomatic bacteriuria and prematurity/low birth weight (LBW) is still a controversial issue, despite many studies. Meta-analysis, a research tool designed to analyze and combine the results of previous studies, may resolve this discrepancy among contradictory results of clinical trials. The purpose of this study was to examine the relationship between asymptomatic bacteriuria and preterm delivery/LBW using meta-analysis. Reports from the literature were classified according to study design into cohort or randomized-treatment control trials. Meta-analysis of cohort studies showed that untreated asymptomatic bacteriuria during pregnancy significantly increased rates of LBW and preterm delivery. Nonbacteriuric patients had only about two-thirds the risk (typical relative risk = 0.65; 95% confidence interval 0.57, 0.74) of LBW and half the risk (typical relative risk = 0.50; 95% confidence interval 0.36, 0.70) of preterm delivery of those with untreated asymptomatic bacteriuria. These reduced risks correspond to a 3.4 (confidence interval 1.8, 5.0) percentage-point difference in LBW and a 3.8 (1.1, 6.4) percentage-point difference in preterm delivery. The analysis of randomized clinical trials showed that antibiotic treatment significantly reduced the risk of LBW (typical relative risk = 0.56; 95% confidence interval 0.43, 0.73), with a substantial reduction of 6.4 (confidence interval 3.3, 9.5) percentage points in the rate of LBW. We conclude that clinical and epidemiologic evidence indicates a strong association between untreated asymptomatic bacteriuria and LBW/preterm delivery and that antibiotic treatment is effective in reducing the occurrence of LBW.