To understand the preventable fraction of low birthweight (LBW) deliveries due to maternal exposure to air pollution during pregnancy in Korea, it is important to quantify the population-attributable risk (PAR). Thus, we investigated the association between maternal exposure to air pollution during pregnancy and LBW, and calculated the PAR for air pollution and LBW in seven Korean cities. We used birth records from the Korean National Birth Register for 2004. A geographic information system and kriging methods were used to construct exposure models. Associations between air pollution and LBW were evaluated using univariable and multivariable logistic regression, and the PAR for LBW due to air pollution was calculated. Of 177 660 full-term singleton births, 1.4% were LBW. When only spatial variation of air pollution was considered in each city, the adjusted odds ratios unit of particulate matter <10 microm in diameter (PM(10)) for LBW were 1.08 [95% confidence interval [CI] 0.99, 1.18] in Seoul, 1.24 [95% CI 1.02, 1.52] in Pusan, 1.19 [95% CI 1.04, 1.37] in Daegu, 1.12 [95% CI 0.98, 1.28] in Incheon, 1.22 [95% CI 0.98, 1.52] in Kwangju, 1.05 [95% CI 1.00, 1.11] in Daejeon and 1.19 [95% CI 1.03, 1.38] in Ulsan. The PARs for LBW attributable to maternal PM(10) exposure during pregnancy were 7%, 19%, 16%, 11%, 18%, 5% and 16% respectively. Because a large proportion of pregnant women in Korea are exposed to PM(10)--which is associated with LBW--a substantial proportion of LBW could be prevented in Korea if air pollution was reduced.