Treatment results (CR/PR** 30-40%) with the combination of 5-FU and FA--a biomodulator of 5-FU action--indicate that FA/5-FU is superior to 5-FU alone in the treatment of advanced colorectal cancer. However the "optimal" dose and time schedule has not been properly defined either for FA or for 5-FU. Moreover, in spite of the promising reports with FA/5-FU, the benefit of chemotherapy in colorectal carcinomas remains unproven. One way to better define the impact on survival and palliation might be to select patients with a documented tumor progression prior to chemotherapy. In 2 prospective phase II studies with FA/5-FU, tumor progression was the essential eligibility criterion. In both studies an overall tumor control (CR/PR/MR/NC)** of progressive disease with relief of tumor-related symptoms in most of the patients was achieved. The median survival times exceeded 12 months. In progressive colorectal cancer these treatment results can be regarded as a significant change in the natural history of the disease. Moreover, such a procedure will help to define more homogeneous patient groups and contribute to a better comparability of different studies. Additionally, overtreatment of patients with slowly proliferating tumors can be avoided, particularly with respect to the subjective and objective side-effects of FA/5-FU combinations.