Chromosome 9 allelic losses have been reported as a frequent and early event occurring in bladder cancer. It has been postulated that a candidate tumor suppressor gene may reside on this chromosome, alterations of which may lead to the development of a subset of superficial bladder tumors. More recently, the involvement of two different regions harboring suppressor loci, one on each of both chromosome 9 arms, has been proposed. We undertook the present study with the objectives of better defining the deleted regions of chromosome 9 in bladder tumors, as well as evaluating the frequency of microsatellite alterations affecting certain loci on this chromosome in urothelial neoplasia. Seventy-three primary bladder tumors were analyzed using a set of highly polymorphic markers, and results were correlated with pathological parameters associated with poor clinical outcome. We observed that, overall, 77% of the tumors studied showed either loss of heterozygosity for one or more chromosome 9 markers and/or microsatellite abnormalities at chromosome 9 loci. Detailed analyses showed that two regions, one on 9p at the interferon cluster, and the other on 9q associated with the q34.1-2 bands, had the highest frequencies of allelic losses. Furthermore, Ta lesions appeared to present mainly with 9q abnormalities, while T1 tumors displayed a mixture of aberrant 9p and 9q genotypes. These observations indicate that loss of heterozygosity of 9p may be associated with the development of superficial tumors with a more aggressive biological behavior or, alternatively, they may be related to early disease progression. In addition, microsatellite alterations were documented in over 40% of amplified cases. Taken together, these data suggest that two different tumor suppressor gene loci on chromosome 9 are involved as tumorigenic events in bladder cancer and that chromosome 9 microsatellite alterations are frequent events occurring in urothelial neoplasia.