The formation of DNA and protein adducts by environmental pollutants is modulated by host polymorphisms in genes that encode metabolizing enzymes. In our study on 67 smokers, aromatic-DNA adduct levels were examined by nuclease P1 enriched 32P-postlabelling in mononuclear blood cells (MNC) and 4-aminobiphenyl-haemoglobin adducts (4-ABP-Hb) by gas chromatography-mass spectroscopy. Genetic polymorphisms in glutathione S-transferase M1 (GSTM1), T1 (GSTT1) and N-acetyl-transferase 1 (NAT1) and 2 (NAT2) were assessed by polymerase chain reaction-based methods. DNA adduct levels, adjusted for the amount of cigarettes smoked per day, were higher in GSTM1(-/-) individuals (1.30 +/- 0.57 adducts per 108 nucleotides) than in GSTM1(+) subjects (1.03 +/- 0.56, P = 0.05), higher in NAT1 slow acetylators (1.58 +/- 0.54) than in NAT1 fast acetylators (1.11 +/- 0.58, P = 0.05) and were also found to be associated with the NAT2 acetylator status (1.29 +/- 0.64 and 1.03 +/- 0.46, respectively, for slow and fast acetylators, P = 0.06). An effect of GSTT1 was only found in combination with the NAT2 genotype; individuals with the GSTT1(-/-) and NAT2-slow genotype contained higher adduct levels (1.80 +/- 0.68) compared to GSTT1(+)/NAT2 fast individuals (0.96 +/- 0.36). Highest DNA adduct levels were observed in slow acetylators for both NAT1 and NAT2 also lacking the GSTM1 gene (2.03 +/- 0.17), and lowest in GSTM1(+) subjects with the fast acetylator genotype for both NAT1 and NAT2 (0.91 +/- 0.45, P = 0.01). No overall effects of genotypes were observed on 4-ABP-Hb levels. However, in subjects smoking less than 25 cigarettes per day, 4-ABP-Hb levels were higher in NAT2 slow acetylators (0.23 +/- 0.10 ng/g Hb) compared to fast acetylators (0.15 +/- 0.07, P = 0.03). These results provide further evidence for the combined effects of genetic polymorphisms in GSTM1, GSTT1, NAT1 and NAT2 on DNA and protein adduct formation in smoking individuals and indicate that, due to the complex carcinogen exposure, simultaneous assessment of multiple genotypes may identify individuals at higher cancer risk.