Integrating systemic chemotherapy into the treatment of patients with invasive bladder cancer, where the majority of deaths are from systemic relapse, is crucial if survival is to be improved. However, the global recommendation of a single treatment approach for all patients is becoming outdated as several groups have found that appropriately selected patients enjoy comparable survival whether treated by transurethral resection alone or by partial or radical cystectomy. Thus, refining case selection becomes a critical area of investigation. Patients with a high risk of systemic relapse should be considered for systemic therapy early in the course of the disease, ideally as part of a clinical trial. The availability of growth factors has reduced the toxicities of the regimens currently in use. Improvements in assessing biologic potential are required, so that treatment recommendations will allow patients the maximal chance of cure and maintenance of organ function, while minimizing toxicities in patients for whom systemic approaches are unwarranted. Of interest are recent reports that G-CSF may enhance tumor sensitivity to methotrexate in vitro and increase the sensitivity of implanted urothelial tumors to chemotherapy in nude mice. Such findings suggest an expanded role for these agents. Unfortunately, simple escalation of all components of a combination regimen does not appear to be a viable strategy, as it is unlikely to significantly increase CR proportions without prohibitive toxicities. However, as more is understood about drug resistance, and in particular its development in vivo, better sequencing of the available of options should be possible. The availability of effective salvage therapies suggests that this is an appropriate therapeutic approach. In addition, a number of strategies aimed at reversing the mdr phenotype are under study. These include calcium channel blockers such as verapamil, cyclosporin, and tamoxifen. Alternatively, some groups are investigating transfecting the mdr gene into bone marrow cells to reduce the sensitivity of these cells to cytotoxic agents. These novel designs can be tested both in patients with metastatic disease and in patients with locally advanced (T3b, T4, and N+) disease, who have a high risk of metastatic failure and low CR proportions to available regimens.