A Nordic multicenter study in asymptomatic patients with advanced colorectal cancer compared initial chemotherapy with sequential methotrexate-5-FU with leucovorin rescue (MFL) for 6 months versus primary expectancy with chemotherapy only after the appearance of symptoms. The study (183 patients randomized between January 1985 and February 1990) showed that symptom-free survival, progression-free survival and survival respectively were about 6 months longer in the group of patients randomized to initial MFL. Whether these prolongations could be achieved without an impaired 'quality of life' was studied in an associated study. Between January 1985 and March 1987, 43 patients were randomized at one of the hospitals, 36 of which were interviewed with a questionnaire at randomization. Even if all these patients were considered, by the physician, to be 'free of symptoms from their disease', 16/36 (44%) had symptoms that could be referred to the disease. In spite of this, the patients were in a good general condition, and considerably better off than patients considered to have 'symptoms from the disease' who were interviewed with the same questionnaire when randomized in a parallel study of symptomatic patients. Patients randomized to initial chemotherapy and interviewed longitudinally maintained their good condition throughout treatment. Toxicity was mild, although the patients expressed more adverse effects than the physicians recorded. Since symptom-free survival, progression-free survival and survival were statistically significantly longer in the group of patients randomized to MFL also in this associated study, it is concluded that initial chemotherapy can prolong symptom-free survival and survival without reduced 'quality of life' during the treatment period.