DNA analysis is becoming an important diagnostic and prognostic adjunct test in urinary cytology. The aim of this study was to compare the results of DNA flow cytometry (FCM) with the cytologic diagnosis of bladder washings (BW). DNA ploidy was evaluated in 251 BW. In 65 cases, follow-up surgical biopsies were available. Cytology results were classified as positive and negative, and FCM results were categorized as diploid and aneuploid. Both tests were evaluated independently. Cases were defined as discordant if the cytology was negative and FCM was aneuploid, or if the cytology was positive and FCM was diploid. All discordant cases were reviewed, and positive predictive values (PPV) for FCM and cytology were calculated for cases with follow-up biopsy results. Cytologic evaluation classified 181 cases as negative, with 175 of them diploid and 6 aneuploid; and 70 as positive, with 53 of them diploid and 17 aneuploid. Overall, there were 59 discordant cases (23.5%, with a confidence limit of 18.2-28.8%). Of 6 aneuploid/ cytology-negative cases, biopsies were available in 4 cases and showed one grade 1, two grade 2, and one grade 3 urothelial carcinoma (UC). Reanalysis of these 6 cytology specimens showed 1 case that should have been interpreted as positive (false negative), 4 true negatives, and 1 polyoma virus infection. Out of 53 diploid/cytology-positive cases, biopsies were available in 45 cases and showed nine grade 1, 14 grade 2, three grade 3 UCs, 11 UCs in situ, and eight negative biopsies. The PPV for cytology was 85%, and the PPV for FCM was 95%. We concluded that FCM, which requires a large number of cells, often cannot detect small aneuploid populations, which are present particularly in cases of UC in situ. Diagn.