We previously reported on the interrelation between tumor stages, tumor grades, and host immunocompetence examined by delayed cutaneous hypersensitivity (DNCB test and PPD test) and the degree of lymphocyte blastogenesis. In this report we have made a study on T-cell distribution in the tissue of bladder carcmoma using immunofluorescent techniques. Antithymocyte serum was obtained by the following method. Human thymus tissue was obtained during an operation, using a fine needle to destroy the tissue. Three million thymocyte cells were injected per a week to the rabbit. After 8 to 10 booster injection, antithymocyte serum of the rabbit was absorbed with human erythrocytes and liver cells. The method used for immunofluorescent staining was that of Saint-Maries with a small modification. The materials consisted of 37 bladder carcinoma made up of 32 cases of transitional cell carcinoma, of squamous cell carcinoma, and 3 of undifferentiated carcinoma. There were 11 cases of stage A, 15 of stage B, 8 of stage C, and 3 of stage D. According to AFIP classification of the grade, they were made up of2 grade 1 17 grade 2, 12 grade 3 and 6 grade 4. Lymhocyte reaction in H. E. stain was examined according to the following criteria. Under the lower power field, a few lymphocytes in the tissue was defined (-); lymphocytes aggregation, (+); one follicular formation of the lymphocytes (++); and more than one (+++). Using an immunofluorescent microscope, T-lymphccyte reaction was examined according to the following criteria. Less than 10 T-lymphocytes in the viewing field was defined (-); 10 to 50 T-lymphocytes (+); 50 to 200 T-lymphocytes (++); and more than 200 T-lymphocytes (+++). The following results were obtained. 1) Interrelation between tumor stages and lymphocyte reaction in the H. E. stained tissue was not significant by the qui-sqare analysis. But interrelation between tumor grades and lymphocyte reaction in the H. E. stained tissue was 0.025 by the qui-sqare analysis. 2) Even in a specimen judged as (+) lymphocyte reaction in H. E. stain, we found the area of light T-Iymphocyte reaction as well as that of heavy T-lymphocyte reaction in the same slice of tissue when observed under the immunofluorescent microscope. 3) Under the immunofluorescent microscope, lymphocyte reaction around the tumor and at the apex of the growing tumor was observed as the reactions predominantly of the T-lymphocyte. 4) T-lymphocyte population was more striking in the low stage cases than in the high stage cases. 5) No clear relationship between T-Iymphocyte population in the tissue and the T-Iymphocyte count of the peripheral blood was observed in the 16 patients studied.