Personal exposure to PM2.5 and PM1, together with indoor and residential outdoor levels, was measured in the general adult population (30 subjects, 23-51 years of age) of Gothenburg, Sweden. Simultaneously, urban background concentrations of PM2.5 were monitored with an EPA WINS impactor. The 24-h samples were gravimetrically analyzed for mass concentration and black smoke (BS) using a smokestain reflectometer. Median levels of PM2.5 were 8.4 mu g/m(3) (personal), 8.6 mu g/m(3) (indoor), 6.4 mu g/m(3) (residential outdoor), and 5.6 mu g/m(3) (urban background). Personal exposure to PM1 was 5.4 mu g/m(3), while PM1 indoor and outdoor levels were 6.2 and 5.2 mu g/m(3), respectively. In non-smokers, personal exposure to PM2.5 was significantly higher than were residential outdoor levels. BS absorption coefficients were fairly similar for all microenvironments (0.4-0.5 10(-5) m(-1)). Personal exposure to particulate matter ( PM) and BS was well correlated with indoor levels, and there was an acceptable agreement between personal exposure and urban background concentrations for PM2.5 and BS2.5 (r(s) = 0.61 and 0.65, respectively). PM1 made up a considerable amount (70-80%) of PM2.5 in all microenvironments. Levels of BS were higher outdoors than indoors and higher during the fall compared with spring. The correlations between particle mass and BS for both PM2.5 vs. BS2.5 and PM1 versus BS1 were weak for all microenvironments including personal exposure. The urban background station provided a good estimate of residential outdoor levels of PM2.5 and BS2.5 within the city (r(s) = 0.90 and 0.77, respectively). Outdoor levels were considerably affected by long-range transported air pollution, which was not found for personal exposure or indoor levels. The within-individual ( day-today) variability dominated for personal exposure to both PM2.5 and BS2.5 in non-smokers.