We sought to determine by what amount the frequency of low birth weight (LBW) could be reduced by eliminating its known, preventable risk factors. Based on the prevalence of risk factors such as cigarette smoking (exposure) in a population, and the degree of association (relative risk) between the risk factors and the outcome of LBW, we estimated the percentage of LBW in the United States attributable to the risk factors (population-attributable risk percentage). We found the following associations with known preventable LBW: Cigarette smoking accounts for 9.6%; alcohol consumption during pregnancy, 2.3%; low pre-pregnancy weight (less than 54 kg), 6.3%; urinary tract infections, 1.3%; Chlamydia trachomatis genital infections, 4.8%; and narcotics addiction, 3.8%. Because these risk factors are not mutually exclusive, eg, narcotics addicts may be addicted to both narcotics and cigarettes, the maximum preventable population-attributable risk percentage for LBW is less than the sum of the individual risk factors, which is 28.1% (9.6% [cigarette smoking] + 2.3% [alcohol consumption] + 6.3% [low pre-pregnancy weight] + 1.3% [urinary tract infection] + 4.8% [Chlamydia] + 3.8% [narcotics addiction] = 28.1%). Given the current state of knowledge about preventable risk factors, LBW can be reduced--in an idealized model--from only 6.9 to 5.0% of all births. In a more realistic model, we estimate that LBW can be reduced from only 6.9 to 6.3% of all births. Therefore, a national effort is indicated to identify unknown preventable causes of LBW.