The clinical usefulness of the nuclear matrix protein 22 (NMP22) Bladder Chek test as a novel urine marker in the detection of patients with bladder cancer was evaluated in comparison with the urinary NMP22 enzyme-linked immunosorbent assay (ELISA) and urinary cytology. A total of 40 patients with pathologically proven bladder cancer voided urine specimen before treatment. The urine samples were divided for NMP22 Bladder Chek test, NMP22 ELISA, and urinary cytology. In the 40 patients with bladder cancer, the overall positive rate was 62.5% for the NMP22 Bladder Chek test, 55% for the NMP22 ELISA test, and 27.5% for urine cytology. There was a significant difference between NMP22 Bladder Chek, NMP22 ELISA and cytology. The positive rate with the NMP22 Bladder Chek and NMP22 ELISA was higher in the patients with high grade and large-size (1 cm < or =) tumor. In 40 patients presenting with microhematuria without urothelial cancer, the false positive rate 12.5, 10, and 0% for NMP22 Bladder Chek, NMP22 ELISA, and urinary cytology. No significant difference was found with the test. In conclusion, the urine NMP22 Bladder Chek test provided a higher positive rate than the NMP22 ELISA test and urinary cytology. Therefore, the NMP22 Bladder Chek test may be clinically more useful as a tumor marker for the diagnosis of bladder cancer.