The Willems badge, a short-term diffusion sampler, was used to measure nitrogen dioxide concentrations inside and outside the homes of participants in the European study "PEACE' (Pollution Effects on Asthmatic Children in Europe). The main aim of the study was to determine levels of nitrogen dioxide concentrations both outside and inside children's homes, and to estimate the indoor/outdoor ratios for nitrogen dioxide in an urban area, in comparison with a less urbanized control area. We conducted measurements in 23 homes in Umeå, a city of about 100,000 inhabitants in the northern part of Sweden, in addition to 20 homes in a less urbanized control area situated about 20 km from Umeå. Measurements were made on two different occasions in each home during the period January-March, 1994. The houses were not equipped with any gas appliances. The mean outdoor 24-h concentration in Umeå was 28 micrograms m-3 and the mean indoor concentration was 11 micrograms m-3. The mean indoor: outdoor ratio was 0.44 (s = 0.23). The highest outdoor value, measured in the city centre of Umeå, was 54 micrograms m-3. In the control area the mean ambient 24-h concentration was 12 micrograms m-3, approximately half as high as in the urban area, and the mean indoor concentration was 6 micrograms m-3. The mean indoor: outdoor ratio was 0.67 (s = 0.55). The correlation coefficient between indoor and outdoor concentrations was higher in the control area, r = 0.79 (p < 0.001), in comparison with the urban area, r = 0.43 (p < 0.01). It is concluded that the outdoor as well as the indoor concentrations of nitrogen dioxide were approximately twice as high in Umeå as in the control area. This could be explained by heavier traffic density in Umeå. The mean 24-h concentration outside homes in Umeå was, however, below the 24-h national standard level of 75 micrograms m-3. The higher correlation between indoor and outdoor concentrations, combined with higher indoor: outdoor ratio, in the control area is interpreted as a sign of a lower level of penetration of outdoor air into the houses in the urban area. This was not explained by differences in types of buildings between the two areas, but possibly by differences in air-exchange rates and in habits of ventilating rooms with open windows.