The aim of this study was to verify a possible correlation between CYP1A1 induction, MspI genotype and lung cancer incidence. A case-control study was performed on 48 lung cancer patients and 81 healthy subjects to test the existence of a correlation, within a European population. The hyperinducible group exhibited a significantly higher risk of lung cancer (odds ratio = 3.41; P = 0.036), especially for adenocarcinoma (odds ratio = 5.29; P = 0.033). In contrast with the situation observed in Asian populations, the frequency of the M2 allele did not differ significantly in the total lung cancer population (7.82%) and the group of healthy subjects (10.71%). The median inducibility value was slightly higher among cancer patients with one or two M2 alleles than among patients homozygous for the wild-type allele (P = 0.09). However, the percentage of individuals possessing at least one mutated allele was not significantly higher among hyperinducible patients (37.5%) than among non-hyperinducible patients (16.0%). No significant correlation could be found between M2 allele and lung cancer or between M2 allele and CYP1A1 inducibility; the only positive correlation found was between CYP1A1 hyperinducibility and lung cancer incidence. Our observations do not support the view that the presence of the M2 allele at the MspI site of the CYP1A1 gene constitutes a significant lung cancer risk in Caucasians.