Epidemiological studies have shown an increased incidence of lung cancer, bladder cancer, and esophageal cancer in chimney sweeps, probably due to their exposure to PAH in soot. The work environment for sweeps has, however, improved during the last decades. It was thus important to assess whether the present exposure still may cause genotoxic effects. A further objective was to assess whether genetic polymorphisms in metabolic enzyme activities could explain some of the variation in the parameters of genotoxicity. Venous blood samples were drawn from 71 chimney sweeps and 59 control subjects. Micronuclei were analyzed in activated peripheral B- and T-lymphocytes with preserved cytoplasm. Polymorphisms for CYP1A1 and GST1 in the sweeps were analyzed by a PCR technique. The sweeps did not have higher frequencies of micronuclei in B- or T-lymphocytes than the control subjects, when allowance was made for age and smoking in a multiple regression analysis. Further, there was no association between years of active work as a sweep and any of the two micronucleus parameters. None of the sweeps had the rare CYP1A1 genotype val/val and only one individual had the m2/m2 genotype. The presence of at least one GST1 allele (GST1+) was observed in 36 subjects (51.4%). Thirteen individuals (18.6%) were of the m1/m2 or m2/m2 genotype. And among those only seven had the combined GST1- and m1/m2 genotype. No difference was observed in B- or T-lymphocyte micronucleus frequencies between sweeps with the rare CYP1A1 genotypes m1/m2, m2/m2 or ile/val compared to individuals with the m1/m1 and ile/ile genotypes. Moreover, the GST1 deficient sweeps (GST1-) did not show any altered micronucleus frequency compared to the GST1 positive sweeps. A possible reason for the lack of genotoxic effect in sweeps is the improved hygienic conditions and change in fuels, which has decreased the exposure levels for PAH. Host polymorphisms for metabolizing enzymes did not influence the micronucleus frequencies. As the sweeps did not differ from the control subjects, with respect to micronucleus frequencies, no conclusion on the importance of host polymorphisms for genotoxic risk can be drawn.