Flow cytometric DNA ploidy analysis has been reported to be more objective and sensitive than morphologic evaluation as a surveillance method in patients with Barrett esophagus (BE) for the development and progression of precancerous lesions. Such analyses are typically performed using fresh samples that require a separate or "jumbo" biopsy, are prone to false DNA aneuploidy if not promptly processed, and do not allow for retrospective studies. The feasibility of performing flow cytometric DNA analysis on paraffin-embedded biopsies was studied to circumvent some of these problems using 12 squamous esophageal mucosa with inflammation and 58 BE cases showing varying degrees of dysplasia. Among the BE cases, 12 had no dysplasia, 20 were indefinite for dysplasia, 14 had low grade dysplasia, and 12 had high grade dysplasia. Satisfactory histograms were obtained in 86% of the analyzed samples. Among cases with adequate histograms, DNA aneuploidy was identified in 77% with high grade dysplasia, 16% with low grade dysplasia, 23% of indefinite for dysplasia, and 0% without dysplasia. One of the esophagitis samples was also DNA aneuploid. Correlation of DNA aneuploidy and degree of dysplasia is highly significant (P = .001). The authors have demonstrated that routinely processed paraffin-embedded biopsies can be used for flow cytometric ploidy analysis. DNA aneuploidy was highly correlated with degree of dysplasia and serves as a quantitative prognostic indicator for prospective as well as retrospective studies of the evolution of BE to carcinoma.