The accumulation of genetic alterations plays a role in the evolution of bladder cancer. These changes can be detected in the urine by DNA analysis of the cells exfoliated from the bladder wall enabling us to detect bladder cancer. The urine supernatant, besides the urine sediment, contains DNA, however in a much smaller amount. The origin of DNA in these two fractions is probably different. Our aim was to evaluate which fraction (supernatant or sediment) provides more reliable results in detecting tumors. We analyzed blood, urine and tumor samples taken from 80 individuals (44 patients with bladder cancer, 20 control patients and 16 healthy volunteers) by using 12 microsatellite markers mapped on 6 chromosomes. Microsatellite alterations were detected in the urine sediment and supernatant in 86% of the cancer cases. Urine sediment alone had a sensitivity of 68%, while urine supernatant alone indicated aberrations in 80% of the tumors. In the superficial (Ta/T1) cases, a considerable difference in sensitivity, 84 vs. 67%, was found between the two fractions in favor of urine supernatant. We also detected deletions in the control groups, although in a much lower proportion. Loss of the 16q24 chromosomal region showed a significant correlation with tumor stage (p=0.02). Microsatellite analysis of the urine is an efficient and noninvasive molecular method to detect bladder cancer. The analysis of free DNA in the urine supernatant provides a higher detection rate. The marker on the chromosomal region 16q24 is suggested to have a prognostic value.