Six bladder cancer patients received intralesional injections by needle, under cystoscopic control, of 1969-4046 units (U) of xenogeneic IL-2 of high biological activity (324 U/ml; 397 micrograms/ml protein; 1.22 protein/U ratio). The treatment was spread over 7-54 days and 0.5 ml was injected each time. In 3/6 patients complete tumour regression was seen 43, 60 and 105 respectively days after the first IL-2 injection. In 2 a 70% regression was observed at days 45 and 75. In the last patient massive necrosis throughout the tumour mass was recorded on day 25 at radical cystectomy. In order to evaluate the minimum IL-2 U required to obtain positive clinical results and/or to assess whether the anti-tumour effect observed could be ascribed to the foreign protein of bovine origin contained in our IL-2 preparation, 4 additional bladder cancer patients were treated in 7-14 days with 156-1404 U of a second IL-2 lot with a much lower biological activity and similar protein content (52 U/ml; 289 micrograms/ml of protein; 5.55 protein/U ratio). No clinical or histological improvement was noted over a 42- to 54-day observation period. When we evaluated the 2 groups of patients by Student's t-test for both total U injected and U/kg of body weight (bw) we found a statistically significant differences (0.0025 less than p less than 0.0005 and p less than 0.0005, respectively). In contrast, no difference was seen for the injected protein amounts. The reported observations are in favour of a dose-dependent anti-tumour action mediated by IL-2 instead of foreign proteins. In none of the patients treated were any early or late adverse clinical side effects observed. Immunological monitoring (E, EAC, E-active rosettes, mitogen lymphocyte stimulations and leukocyte migration inhibition in the presence of allogeneic bladder cancer cells) performed on the peripheral blood (PB) showed significant but contrasting modifications after IL-2 injection. There was no clear correlation with the clinical course. The patients in whom we observed complete regression are still tumour free after 2, 4 and 7 months. In addition, in all the patients of the first group we observed an increase in tumour lymphoid infiltrate after IL-2 injection and in 2 patients lymphoid pseudo-follicles were also noted. In 2 of these patients we also observed scar-like areas in the place of the tumours previously seen.