Chemotherapy treatments for urothelial tract tumors have improved to the point that some patients are enjoying long-term disease-free survival. Moreover, with currently available agents and combinations, and with our increased application of clinical and biologic prognostic factors, we are refining our ability to select appropriate therapies for individual patients. We have learned that once the decision is made to use combination chemotherapy, adequate doses should be used. This can be facilitated by the coadministration of hematopoeitic growth factors. Recently completed phase II trials have confirmed that higher doses and dose rates may increase response proportions of and in particular, the proportion of complete responses. The finding that granulocyte colony stimulating factor enhances the sensitivity of tumor cells to methotrexate in vitro and to other agents studied against urothelial tumors implanted in nude mice implies an expanded role for these compounds. However, because non-hematologic toxicities are still important, it is unlikely that simple escalation of all components a four drug regimen such as of M-VAC (cisplatin, methotrexate, vinblastine, and doxorubicin) will have a significant impact on survival. In addition, as more is learned about the pharmacokinetic and pharmacodynamic relationships of the active agents, it appears that better schedules can be designed to improve the therapeutic index of the compounds. Ultimately we will be able to determine drug sensitivities, both at the start of therapy and as it evolves during treatment, that will allow a better selection of a particular chemotherapeutic regimen. For example, mdr1 induction appears to play a significant role in the therapy for treatment-resistant tumors. The availability of a number of active salvage regimens that are not constrained by this mechanism hints that changes in drug sequencing and drug scheduling may provide a significant improvement in outcome. While established combination chemotherapy regimens should be considered standard therapy in appropriately selected patients, promising strategies and new agents need to be investigated if we are to "break the deadlock" that has appeared in the treatment of urothelial tumors. These investigations can be performed safely in a well-controlled fashion to enable the identification of new regimens and to compare promising strategies with appropriate control populations in randomized trials.