A method for measuring personal nitrogen dioxide (NO2) exposure, using passive samplers, was tested among schoolchildren. Activity patterns and NO2 exposure levels were studied in relation to urban and rural living. Stationary air monitoring data indicated that the urban children were supposed to be exposed to NO2 levels that were among the highest in Sweden. It was shown that NO2 levels measured at the stationary air monitoring station were not representative for the children's exposure. The children spent 90% of their time indoors; only a small percentage of their time was spent in transit. The median daily NO2 exposure level in the urban area (13 micrograms NO2/m3, 7 ppb) was significantly higher (p < .001) than in the rural area (7 micrograms NO2/m3, 4 ppb). The most important source of exposure was the indoor ice-skating arenas, where levels up to 8,000 micrograms NO2/m3 (4 240 ppb) were measured during 1-h periods.