Summary We have used a relatively new trial methodology, the group sequential design, to prospectively evaluate two dose levels of bolus/infusional 5-fluorouracil (5-FU) and folinic acid in 192 consecutive-patients with advanced colorectal carcinoma. On day 1, all patients received 200 mg m-2 of folinic acid infusion over 2 h. Cohort A (n = 102 patients) received 500 mg m-2 5-FU by i.v. 1 5-min infusion followed by an infusion of 500 mg m-2 5-FU over 22 h. Treatment was repeated on day 2 and further cycles given 2-weekly. After sequential analysis excluded a response rate of over 40%, cohort B (n = 90 patients) received an increased dose of 600 mg m-2 5-FU bolus and infusion. Patients had received no prior 5-FU therapy and the two cohorts had similar demographic features. In 179 evaluable patients, the overall response rate was 18% (95% Cl 12-24%) with CR of 6% and PR of 12%, with no difference between the two cohorts. Overall median survival was 34 weeks (95% Cl 30-39) with no significant difference between cohorts (median survival 32 and 37 weeks in cohort A and B respectively; P = 0.27). On multivariate analysis, poor performance status, elevated initial WBC and alkaline phosphatase and low serum albumin were associated with reduced survival (P < 0.05), and initial raised WBC showed an association with reduced likelihood of response (P = 0.002). Overall toxicity was low with CTC grade 3 mucositis, diarrhoea, nausea or vomiting in < 6% of patients and no treatment-related deaths. Significant (grade 3 or above) leucopenia was more common in cohort B than in cohort A (9% and 1% respectively); there were more dose reductions, and the median administered dose intensity was lower in cohort B than in cohort A (89% and 97% respectively; P = 0.006). In this group of relatively unselected patients, we have confirmed a relatively low objective response rate and median survival of 7.8 months with this regimen. There was no significant difference in outcome between the two dose levels but the higher dose of 5-FU was associated with more dose reductions and greater toxicity.Keywords: 5-FU; folinic acid; metastatic colorectal carcinoma; group sequential triangular test Colorectal cancer is second only to bronchial carcinoma as a cause of cancer death in the UK. There are currently 28 000 new cases in the UK each year, and around half of these patients will die within 5 years of diagnosis. A meta-analysis of 5-fluorouracil (5-FU) used alone in a conventional bolus schedule showed a disappointing response rate of 11 % (Advanced Colorectal Cancer Metaanalysis Project, 1992).A major locus of action of 5-FU is the enzyme thymidylate synthase (TS), which is responsible for the formation of dTMP from dUMP. The 5-FU metabolite FdUMP is a potent inhibitor of TS, forming an irreversible ternary complex with TS and the cofactor 5,10-CH2-tetrahydrofolate (5,10-CH2-FH4). Reduced intracellular concentrations of 5,10-CH2-FH4 may therefore limit the formation of the ternary complex and hence limit cytotoxicity of 5-FU. This hypothesis prov...