We reported previously that increases in ambient air pollution in the Los Angeles basin increased the risk of low weight and premature birth. However, ambient concentrations measured at monitoring stations may not take into account differential exposure to pollutants found in elevated concentrations near heavy-traffic roadways. Therefore, we used an epidemiologic case-control study design to examine whether residential proximity to heavy-traffic roadways influenced the occurrence of low birth weight (LBW) and/or preterm birth in Los Angeles County between 1994 and 1996. We mapped subject home locations at birth and estimated exposure to traffic-related air pollution using a distance-weighted traffic density (DWTD) measure. This measure takes into account residential proximity to and level of traffic on roadways surrounding homes. We calculated odds ratios (ORs) and risk ratios (RRs) for being LBW and/or preterm per quintile of DWTD. The clearest exposure-response pattern was observed for preterm birth, with an RR of 1.08 [95% confidence interval (CI), 1.01-1.15] for infants in the highest DWTD quintile. Although higher risks were observed for LBW infants, exposure-response relations were less consistent. Examining the influence of season, we found elevated risks primarily for women whose third trimester fell during fall/winter months (OR(term LBW) = 1.39; 95% CI, 1.16-1.67; OR(preterm and LBW) = 1.24; 95% CI = 1.03-1.48; RR(all preterm) = 1.15; 95% CI, 1.05-1.26), and exposure-response relations were stronger for all outcomes. This result is consistent with elevated pollution in proximity to sources during more stagnant air conditions present in winter months. Our previous research and these latest results suggest exposure to traffic-related pollutants may be important.