Cellular glutathione (GSH) levels were measured from 27 human lung tumor biopsies, enzymatically disaggregated, and compared with cells isolated from normal lung of the same patients. GSH levels from normal lung were similar among patients with a mean value of 11.20 +/- 0.58 (SEM) nmol GSH/mg protein (24 patients) with a range from 6.1 to 17.5 nmol GSH/mg protein. GSH levels varied considerably within and across histological tumor types with the following values: adenocarcinomas, 8.83 +/- 0.96 nmol/mg protein (8 patients); large cell carcinomas, 8.25 +/- 2.51 nmol/mg protein (3 patients); and squamous cell carcinomas, 23.25 +/- 5.99 nmol/mg protein (8 patients). The cyclic GSH reductase assay gave only average GSH values and could not distinguish possible GSH variation among subpopulations of cells isolated. Cell volume measurements and microscopic evaluation of cells isolated from both tumors and normal lung revealed heterogeneity with respect to cell types present. To determine the extent of thiol variation among tumor cell subpopulations, tumor cell suspensions were stained with the thiol-specific stain, monochlorobimane (MCB). The accuracy of MCB staining was tested by flow cytometric analysis of 12 in vitro human tumor cell lines and 3 rodent cell lines. A linear relationship was found between the bimane cellular fluorescence and the cyclic GSH reductase assay for cell lines having less than 80 nmol GSH/mg protein (R2 = 0.82). Above 80 nmol GSH/mg protein the rate of change of the bimane fluorescence intensity with respect to increasing GSH concentrations was much reduced. However, by labeling cells with MCB it was possible to distinguish between cell lines with low versus high GSH content. MCB staining of tumor samples revealed multiple populations of cells with respect to thiol levels. In particular, 2 of 8 squamous cell carcinomas had a proportion of cells with elevated fluorescence intensities (from 10 to 35% of the population) suggesting the presence of cells with greatly elevated thiol levels. These findings underscore the complexity of quantitating intracellular GSH levels from tumor biopsies. The combined use of MCB with flow cytometry and conventional GSH assays may help to delineate subpopulations of cells within tumors with different thiol levels.