To investigate the validity of outdoor particulate matter with a 50% cutoff diameter of 10-microm (PM10) concentrations as a measure of exposure in time series studies, the association between personal and outdoor concentrations, within subjects, over time was investigated. Repeated measurements of personal, indoor, and outdoor PM10 were conducted among 37 nonsmoking, 50- to 70-year-old adults, living in Amsterdam, Netherlands, 1994. Regression analyses were conducted for each subject separately, and the distribution of the individual regression and correlation coefficients was investigated. Furthermore, the extent to which differences among personal, indoor, and outdoor concentrations could be explained was studied. The median Pearson's R between personal and outdoor concentrations was 0.50. Excluding days with exposure to environmental tobacco smoke (ETS) improved the correlation to a median R of 0.71. The estimated cross-sectional correlations were lower, 0.34 and 0.50, respectively. Outdoor concentrations (mean, 42 microg/m3) exceeded indoor concentrations (mean, 35 microg/m3) but underestimated personal exposures (mean, 62 microg/m3). The major part of the difference between personal and outdoor concentrations could be attributed to exposure to ETS, living along a busy road, and time spent in a vehicle. The results show a reasonably high correlation between personal and outdoor PM10 within individuals, providing support for the use of ambient PM10 concentrations as a measure of exposure in epidemiologic studies linking the day-to-day variation in particulate matter air pollution to the day-to-day variation in health endpoints such as mortality, hospital admissions, respiratory symptoms, and lung function.