In 1988, Shafik et al. presented their first attempt to apply cytostatics in the treatment of pelvic malignancies. Experimentally and clinically, they studied the anal route for administration of methotrexate, and their first results showed that higher concentration of drugs can be achieved in the bladder wall compared to that in serum. Since December 1988, we have been engaged in experimental and clinical investigations of the possibilities this method might reveal. Experiments with animals have proven that submucosal injection of the cytostatic agent does not induce local changes. The clinical trial on 20 patients with tumors of stage T2-T4 has not shown side effects, either local or systemic. The treatment was well tolerated by the patients. Anorexia has been reported in 3, elevated temperature in 2 and leukopenia in 4 of 20 patients. Complete response has been documented in 3 (15%), partial response in 3 (15%), stabilization in 5 (25%) and progression of the disease in 9 (45%) of the patients. Mean duration of remissions was 10.3 months (range 9-12). The distribution of submucously injected contrast medium into the anorectal area has been studied by computerized tomography in 4 of the patients. Although our results do not support the original estimates of the inventors of the method, its ease of application, the possibility of out-patient administration and negligible side effects are sufficient arguments for performing further experimental and clinical studies.