The outcome of 336 unselected patients diagnosed as having bladder cancer in 1985 in a southern health region of Norway was studied. Two hundred and forty patients had superficial bladder cancer (Tis, Ta and T1). Seventy-four had T2-3 and 17 had T4 bladder tumours at the time of diagnosis (the T-category was unknown in five cases). In 46 of 248 evaluable cases (19%) 12 or more months had elapsed between the onset of symptoms and the histological confirmation of the diagnosis. The information received from the initial routine histology report was inadequate in 51 of 240 (21%) of the patients with superficial bladder cancer. Among the 91 patients with muscle-infiltrating tumours the primary treatment varied considerably, and only 15 patients underwent total cystectomy as the initial treatment. Only 46 in whom muscle-infiltrating tumours were diagnosed initially were referred to the regional uro-oncological unit during the course of the disease. The cancer-corrected, four-year survival was 86% and 42% for superficial and muscle infiltrating bladder cancer, respectively. The comparable figures for crude survival were 64% and 34%, respectively. The lack of optimal standard treatment of muscle-infiltrating bladder cancer warrants the introduction of clinical trials to assess both curative and palliative regimens as well as to study prognostic factors such as proliferation and immunohistochemical parameters by uro-oncological units. Scandinavian Cancer Registries should consider the optional recording of the T category on the case record forms for newly diagnosed cases of bladder cancer.