The true endpoints of cancer chemotherapy have to be prolongation of survival, palliation of symptoms, or improvement of quality of life in cancer patients, while the response rate is considered to be a surrogate endpoint. When conducting a phase III trial, not only the response rate but also data on survivals in the phase II study should be taken into consideration. New response evaluation criteria and various objective markers in analyzing survival have been developed recently. Our data on chemotherapy for gastrointestinal malignancies reveal correlations between CR cases and survival prolongation among esophageal cancer patients treated with chemoradiotherapy, and between responder and survival prolongation in gastric cancer patients treated with chemotherapy. A review of the literature might reveal a correlation between response and survival prolongation in colorectal chemotherapy patients, though our data did not support such a correlation. At present, it is necessary to recognize that "evidence" should be based on the results of randomized phase III trials with large sample sizes and good quality assurance other than the response rates in phase II trials.