The present study has the aim of evaluating gene-environment interaction on the levels of different biomarkers in coke-oven workers exposed to PAH. In order to assess whether the levels of some biomarkers (PAH-DNA adducts, nitro-PAH adducts to Hb and MN frequency) could be modulated by the genetic metabolic polymorphisms for CYP1A1 and GSTM1, we analysed in 76 coke-oven workers and 18 controls the CYP1A1 (MspI and Ile/Val sites) and the GSTM1 genotypes by a PCR assay. In individuals with shared setup of CYP1A1 or GSTM1 genotypes, we analysed how the specified biomarkers correlated with total PAH exposure (urinary levels of 1-hydroxypyrene) both by a stratified analysis and logistic regression modelling. Statistically significant (P = 0.03 and P = 0.01) higher percentages of the more susceptible GSTM1- subjects compared to the GSTM1+ subjects and of the more susceptible CYP1A1 Ile/Val individuals compared to the CYP1A1 Ile/Ile individuals were detected for high levels of PAH-DNA adducts in the high exposure group (namely high levels of 1-OHP). A statistically significant association was observed between increased PAH-DNA adduct levels and the more susceptible GSTM1- genotype (P.O.R. = 4.18, P = 0.03) in a logistic regression modelling and a significant interaction between PAH exposure and GSTM1-genotype was found for PAH-DNA adducts. No effect of these metabolic genotypes was observed for MN frequency and nitro-PAH adducts to Hb. In conclusion, a gene-environment interaction between PAH exposure and two metabolic genotypes involved in activation (CYP1A1) and detoxification (GSTM1) of PAHs, respectively, has been identified.