A new endpoint, the Clinical Benefit Response (CBR), was recently suggested as a means of measuring the effectiveness of palliative chemotherapy in patients with pancreatic cancer. We took advantage of extensive recordings of measurements of treatment effects in two trials in patients with gastrointestinal cancer in order to assess the relevance of CBR criteria. In 151 patients, 53 with pancreatic cancer, 37 with biliary cancer and 61 with gastric cancer, who were randomized between chemotherapy with best supportive care and best supportive care alone, CBR criteria were applied retrospectively to the evaluations of the effects on the patients' well-being made by the treating physician and by the patient in quality-of-life questionnaires, chiefly EORTC QLQ C-30. When compared with subjective response evaluations made by both the physician and the patients' questionnaire data, it was found that CBR overestimated the beneficial effects in certain patients and underestimated them in others. The reason for this overestimation was mainly that the adverse effects of the chemotherapy were not considered appropriately. The underestimation was mainly due to the fact that CBR in practice is dominated by pain assessment, and alterations in other symptoms are also clinically relevant. CBR did, however, show the same differences between the two randomized groups as the subjective response evaluations and the quality-of-life estimations. It is concluded that CBR can be used as a method of measuring the beneficial clinical impact of a treatment strategy in clinical trials, but its limitations in reflecting several important aspects influencing the overall well-being of patients do, however, restrict its use. The criteria designed for use in patients with pancreatic cancer cannot be immediately applied to patients with other cancer types.