To assess the accuracy of the bronchoscopic DNA content analysis, samples of non-small-cell lung carcinomas (NSCLC) were investigated by means of flow cytometry. Samples were dissociated using the detergent Triton X-100. In 58 NSCLC cases, 39 (67%) had DNA aneuploid tumors. We compared the DNA indices of bronchoscopic brushing samples with 21 corresponding surgical samples. In 16 (76%) cases, DNA ploidy of both bronchoscopic and surgical samples were in concordance. In 3 (14%) cases, both bronchoscopic and surgical sample showed DNA aneuploidy, but the number of the DNA aneuploid stem cell lines was different. The cause of these differences was ascribed to the intratumor DNA heterogeneity. In 2 (10%) cases, the bronchoscopic sample showed DNA diploidy, but the surgical sample showed DNA aneuploidy. In these cases, tumor cells obtained by bronchoscopic brushing were so few that the small DNA aneuploid peak was undetectable in the DNA histogram. But the tumor DNA ploidy was evaluated correctly in 90% of 21 cases using bronchoscopic samples. Consequently, despite some drawbacks, the DNA ploidy diagnosis using bronchoscopic samples in this relatively small study, was almost as reliable as surgical samples.