Effects of low-level air pollution were studied in one polluted city and two reference cities in northern Finland by comparing the frequency of upper respiratory infections over a 12-month period in 1982 as reported by parents of children ages 14 through 18 months (n = 679) and 6 years (n = 759). A similar comparison was carried out between children living in the more polluted and less polluted areas of the polluted city. The annual mean and the greatest half-hour concentrations of sulfur dioxide (23 and 807 micrograms/m3), particulates (31 and 291 micrograms/m3), nitrogen oxides (15 and 160 micrograms/m3), and hydrogen sulfide (2 and 177 micrograms/m3) in the polluted city were mainly due to industrial sources. In the reference cities, air pollution was produced mainly by traffic and heating. Adjusted odds ratios (OR) for one or more upper respiratory infections of residents in the polluted city vs those in the reference cities were 2.0 (95% confidence interval [CI] = 1.3-3.2) in the younger age group and 1.6 (95% CI = 1.1-2.1) in the older age group. Within the polluted city, OR calculated for living in more vs less polluted areas were 2.0 (95% CI = 1.0-4.0) in the younger and 1.6 (95% CI = 1.0-2.7) in the older children. The present results suggests that, for children, air pollution can be hazardous in concentrations lower than those recorded in earlier studies from Britain and central Europe. The synergistic effect of sulfur dioxide, particulates, nitrogen oxides, hydrogen sulfide, and other pollutants may be a contributing factor.