Glutathione S-transferase class mu (GSTM1) is known to detoxify certain carcinogens or their activated metabolites. In a previous study using phenotyping methods, individuals genetically devoid of this enzyme activity were significantly overrepresented among lung cancer patients compared to controls, suggesting that this trait is a risk factor for lung cancer. Here, GST class mu status has been determined both pheno- and genotypically, i.e., (a) by ex vivo measurement of trans-stilbene oxide conjugation in lymphocytes, (b) by GSTM1 quantification in blood using an immunoassay, and (c) by the application of polymerase chain reaction to genomic DNA with characterization of an inactivating mutation responsible for the null allele and a G/C single base allelic variance corresponding to the polymorphism of GSTM1 isoenzymes mu and psi, respectively. One hundred seventeen lung cancer patients and 155 control patients were studied. The two groups were of German origin and were similar with respect to age, sex, smoking history, and catchment area. In about 97% of cases, the three methods of assigning activity type of GSTM1 gave corresponding results. By genotype, 55 of 117 lung cancer patients (47.0%) and 73 of 155 control patients (47.1%) were GSTM1 active. The control group was confirmed by analysis of GSTM1 genotype in 200 further, independently studied reference patients; 101 of them were GSTM1 active (50.5%). Thus, the hypothesis of heritable GSTM1 deficiency as a host factor predisposing to lung cancer proved inappropriate. Detailed analysis of subgroups with respect to smoking habits, age, and sex failed to reveal an impact of GST class mu genotype on lung cancer risk. Among the total of 272 patients studied, 36 individuals carried at least one psi allele; however, no unexpected frequency distribution was observed.