The nuclear DNA content of archival paraffin-embedded bladder cancer samples (70 patients) of WHO grades I-III has been measured by flow cytometry. The female/male ratio was 15/55. The mean follow-up time was 13 years (range 9.6-22.0 years). 37 of 70 (53%) patients had DNA index 1.0 (diploid DNA content), and the remaining 33 (47%) patients had an aneuploid tumor. There was no significant difference in the age (mean +/- SD) of the patients having a diploid (66 +/- 9 years) or an aneuploid tumor (68 +/- 11 years) at the time of diagnosis. 47 deaths occurred during the follow-up period; 24 (51%) of these were due to bladder cancer (12 diploid, 12 aneuploid tumors). No significant difference was found after radical treatment during the disease-free interval (mean +/- SD) between diploid (48 +/- 45 months) and aneuploid (35.5 +/- 35 months) groups of patients. Recurrences during the follow-up period were equally common among aneuploid and diploid tumors. A statistically significant relation between histological grade and survival could be demonstrated, but DNA ploidy and S phase fraction had little prognostic value in this respect. There was no statistically significant difference in survival between aneuploid (30%) and diploid (35%) groups of tumors during the follow-up period. The study suggests that flow cytometric determination of nuclear DNA ploidy from paraffin-embedded samples in bladder tumors does not add to the prognostic power of subjective histological grading.