Ambient particulate air pollution assessed as outdoor concentrations of particulate matter < or = 2.5 microm in diameter (PM(2.5)) has been associated with an increased cancer risk. However, outdoor PM(2.5) concentrations may not be the best measure of the individual particle exposure that is a sum of many sources besides outdoor particle levels, e.g., environmental tobacco smoke and cooking. We measured personal PM(2.5) and black smoke exposure in 50 students four times over 1 year and analyzed for biomarkers of different types of DNA damages. Ambient PM(2.5) concentrations were also measured. Exposure was measured for 48 h, after which blood samples were collected and analyzed for DNA damage in lymphocytes in terms of 7-hydro-8-oxo-2'-deoxyguanosine (8-oxodG), strand breaks, endonuclease III- and fapyguanine glycosylase-sensitive sites, and polyaromatic hydrocarbon adducts. Twenty-four-h urine collections were analyzed for 8-oxodG and 1-hydroxypyrene. Personal PM(2.5) exposure was found to be a predictor of 8-oxodG in lymphocyte DNA with an 11% increase in 8-oxodG/10 microg/m(3) increase in personal PM(2.5) exposure (P = 0.007). No other associations between exposure markers and biomarkers could be distinguished. The genotype of glutathione S-transferase M1 (GSTM1), T1 (GSTT1), and P1 (GSTP1) and NADPH:quinone reductase was also determined, but there were no effects of genotype on DNA polyaromatic hydrocarbon adducts or oxidative damage. The results suggest that moderate exposure to concentrations of PM can induce oxidative DNA damage and that personal PM(2.5) exposure is more important in this aspect than is ambient PM(2.5) background concentration.