Liver metastases of colorectal origin arise from malignant cells entering the portal venous circulation. Four hundred sixty patients from 7 participating institutions have been entered in a prospective randomized trial to assess the value of adjuvant portal venous infusion to prevent liver metastases following curative colorectal cancer surgery. By preoperative randomization, patients were assigned to surgery alone (control arm) or to portal liver infusion with 5‐fluorouracil (5‐FU) (500 mg/m2 daily × 7, continuous infusion during the first 7 postoperative days) and mitomycin C (10 mg/m2, 24 hours postoperatively as a 2‐hour infusion). A portal venous catheter was placed through any side branch of the mesenteric venous system during laparotomy for the primary tumor. Using the transabdominal route, there have been no catheter‐related complications. Despite a large cumulative dose of 5‐FU given during the immediate postoperative period, the systemic side effects were minimal. Overall hospital mortality in this study was 1.85% and was not influenced by the adjuvant treatment. This rate is considerably lower than that reported by previous multicenter trials and by the surgical literature, and it indicates an advance in surgical technique and pre‐/postoperative patient management in this type of elective surgery.Three hundred seventy‐eight patients (187 controls, 191 infusion patients) are completely evaluable for recurrence and survival. After a median follow‐up of 24 months, 43 recurrences have been observed in the control group and 34 in the infusion group (p<0.05). Liver metastases were present in 18 control patients and in 14 infusion patients. Twenty‐five patients in the control group and 18 patients in the infusion group died of recurrent disease. Due to the low number of recurrences in this study, it is too early to draw survival conclusions.Several controlled trials using adjuvant portal infusion are in progress. They also showed the feasibility of this approach and they suggested that adjuvant cytotoxic liver infusion may reduce the incidence of liver metastases without significantly increasing morbidity and mortality. It is too early, however, to recommend adjuvant portal infusion outside a clinical trial setting and such treatment should still be restricted to well‐designed prospective protocols.