Prior studies reported an association between ambient air concentrations of total suspended particles and SO2 during pregnancy and adverse pregnancy outcomes. We examined the possible impact of particulate matter up to 10 microm (PM10) and up to 2.5 microm (PM2. 5) in size on intrauterine growth retardation (IUGR) risk in a highly polluted area of Northern Bohemia (Teplice District). The study group includes all singleton full-term births of European origin over a 2-year period in the Teplice District. Information on reproductive history, health, and lifestyle was obtained from maternal questionnaires. The mean concentrations of pollutants for each month of gestation were calculated using continuous monitoring data. Three intervals (low, medium, and high) were constructed for each pollutant (tertiles). Odds ratios (ORs) for IUGR for PM10 and PM2.5 levels were generated using logistic regression for each month of gestation after adjustment for potential confounding factors. Adjusted ORs for IUGR related to ambient PM10 levels in the first gestational month increased along the concentration intervals: medium 1.62 [95% confidence interval (CI), 1.07-2.46], high 2.64 (CI, 1.48-4.71). ORs for PM2.5 were 1.26 (CI, 0.81-1.95) and 2.11 (CI, 1. 20-3.70), respectively. No other associations of IUGR risk with particulate matter were found. Influence of particles or other associated air pollutants on fetal growth in early gestation is one of several possible explanations of these results. Timing of this effect is compatible with a current hypothesis of IUGR pathogenesis. Seasonal factors, one of the other possible explanations, is less probable. More investigation is required to examine these findings and alternative explanations.