The aim of this study was to evaluate the UroVysion (Vysis, Downers Grove, IL) fluorescence in situ hybridization (FISH) test for improved detection of bladder cancer in urinary specimens. Three groups of specimens were examined, including voided urine specimens (1) collected before resection of bladder cancer, (2) from cystoscopically negative bladders of patients with previous bladder cancer, and (3) from patients with benign prostatic hyperplasia (controls). FISH positivity was defined as more than 2 urothelial cells with an abnormal signal copy number of at least 1 of the 4 probes. FISH was positive in 1 of 27 control specimens and in 33 (73%) of 45 pTa, 12 (100%) of 12 pT1, and 13 (100%) of 13 pT2-4 tumors. The results were similar in a series of 68 bladder washings. In addition, FISH of voided urine specimens was positive in 5 of 10 patients with negative follow-up cystoscopy results. Subsequent recurrence was found in 4 of these patients but in none of 5 patients with FISH-negative results. Multiprobe FISH markedly improves the sensitivity and specificity of cytology for the detection of bladder cancer in urine specimens.