A group of eczema patients topically treated with coal tar (CT) ointments was used as a model population to examine the applicability of DNA adducts in WBC subpopulations as a measure of dermal exposure to polycyclic aromatic hydrocarbons (PAHs). Aromatic DNA adducts were examined by 32P-postlabeling in exposed skin and WBC subsets, and urinary excretion of PAH metabolites was determined to assess the whole-body burden. The median urinary excretion of 1-hydroxypyrene and 3-hydroxybenzo(a)pyrene was 0.39 (range, 0.12-1.57 micromol/mol creatinine) and 0.01 micromol/mol creatinine (range, <0.01-0.04 micromol/mol creatinine), respectively, before the dermal application of CT ointments. After treatment for 1 week, these levels increased to 139.7 (range, 26.0-510.5 micromol/mol creatinine) and 1.18 micromol/mol creatinine (range, <0.01-2.14 micromol/mol creatinine), respectively, indicating that considerable amounts of PAHs were absorbed. Median aromatic DNA adduct levels were significantly increased in skin from 2.9 adducts/10(8) nucleotides (nt; range, 0.7-10.0 adducts/10(8) nt) before treatment to 63.3 adducts/10(8) nt (range, 10.9-276.2 adducts/10(8) nt) after treatment with CT, in monocytes from 0.28 (range, 0.25-0.81 adducts/10(8) nt) to 0.86 adducts/10(8) nt (range, 0.56-1.90 adducts/10(8) nt), in lymphocytes from 0.33 (range, 0.25-0.89 adducts/10(8) nt) to 0.89 adducts/10(8) nt (range, 0.25-3.01 adducts/10(8) nt), and in granulocytes from 0.28 (range, 0.25-0.67 adducts/10(8) nt) to 0.54 adducts/10(8) nt (range, 0.25-1.58 adducts/10(8) nt). A week after stopping the CT treatment, the DNA adduct levels in monocytes and granulocytes were reduced to 0.38 (range, 0.25-0.71 adducts/10(8) nt) and 0.38 adducts/10(8) nt (range, 0.25-1.01 adducts/10(8) nt), respectively, whereas the adduct levels in lymphocytes remained enhanced [1.59 adducts/10(8) nt (range, 0.25-2.40 adducts/10(8) nt)]. Although the adduct profiles in skin and WBC subsets were not identical, and the adduct levels in WBCs were significantly lower as compared with those in skin, the total DNA adduct levels in skin correlated significantly with the adduct levels in monocytes and lymphocytes, but not with those in granulocytes. Excretion of urinary metabolites during the first week of treatment was correlated with the percentage of the skin surface treated with CT ointment and decreased to background levels within a week after the cessation of treatment. 3-Hydroxybenzo(a)pyrene excretion, but not that of 1-hydroxypyrene, correlated significantly with the levels of DNA adducts in skin that comigrated with benzo(a)pyrene-diol-epoxide-DNA. This study indicates that the DNA adduct levels in mononuclear WBCs can possibly be used as a surrogate for skin DNA after dermal exposure to PAHs.