Summary Bladder cancer is the second most common genitourinary tumour and is a significant cause of morbidity and mortality. Trials of neoadjuvant and adjuvant chemotherapy have failed to show a survival advantage, although these studies generally had suboptimum design and an insufficient number of patients. Despite the introduction of newer agents, the median survival for metastatic disease is about 1 year; however, improvements in quality of life have been achieved. Platinum drugs should be included in studies of combination chemotherapy regimens wherever possible. There have been various studies exploring the role of taxanes, gemcitabine, ifosfamide, and platinum in double and triple combinations in different schedules to maximise dose intensity and improve effectiveness but large phase III trials are needed. The current tumour, node, and metastasis staging system is insufficient to predict outcome in patients with bladder cancer irrespective of the treatment they received. Evaluation of molecular prognostic markers should be incorporated into phase II and III trials to define their roles in clinical outcome. Future studies should stratify patients according to the number of risk factors they have to avoid imbalance in treatment groups and patients should be carefully selected.