Urothelial neoplasia is a complex subject that can be only partially understood by careful study of the light microscopic features of individual lesions. Despite decades of study, our knowledge concerning the interaction of neoplastic events and host responses remains rudimentary. Most information has been collected by observing cases (usually in retrospect) that have been grouped according to relatively arbitrary criteria based on the phenotypic appearances of lesions as they are viewed through the light microscope. When evaluating human disease, we are always forced to reason backwards to determine the most likely histogenesis and to project forward to assess the most likely prognosis. Both types of reasoning must be filled with conjecture since direct observations from the initial events to the end results are not possible. Under these constricting circumstances, a conceptual framework into which our often anecdotal observations can be placed is more important than we would like to admit. With this in mind, I have taken advantage of the monograph format to risk a speculative approach to the subject, at least as it applies to the significance of the pathological features. In the past, we have accepted the view that all human hosts are essentially the same and that variations in cancer type and behavior are related almost entirely to the genetic ingenuity of the cancer cells themselves. Perhaps we should now entertain the opposite view, that carcinogenic events are ubiquitous and that cancer in any individual patient represents only what that patient will allow to grow in his or her body. We have devoted almost all of our collective research energy to examining the tumor. I believe that the pace of future progress can be significantly increased if we can think of more ways to examine the patient.