Fifty-nine patients with a history of bladder cancer were evaluated with three serially voided urines for flow cytometry and cytology within the 24 hours preceding cystoscopy, bladder wash, and biopsy. Evaluation of the 32 patients with biopsy proven cancer revealed that for one, two, and three voided urine specimens, the sensitivity of flow cytometry in detecting bladder cancer was 29%, 35%, and 41%, respectively, whereas the sensitivity of voided urine cytology was 44%, 53%, and 57%, respectively. Bladder wash flow cytometry had a sensitivity of 76%. Interpretation of the voided urine flow cytometry was based on a control group of 80 volunteers with negative voided urine cytology. The control group demonstrated that a minimum of 1500 analyzable cells was necessary for a reliable histogram. Considering all 59 patients, voided urine flow cytometry, bladder wash flow cytometry, and voided urine cytology were acceptable for analysis in 53%, 93%, and 100% of the specimens, respectively. Voided urine specimens appear less suitable than bladder irrigation specimens for evaluation by flow cytometry. Voided urine flow cytometry is less sensitive in detecting bladder cancer than voided urinary cytology or bladder wash flow cytometry and is not effective for use as a bladder cancer screening test.