Animal experiments in BALB/c-mice and in DBA/2-mice confirmed that lectin blockade with D-galactose containing receptor analogues can inhibit metastatic spread into the liver. The number of liver colonies of inoculated tumor cells was significantly reduced after D-galactose treatment as compared to animals of control group. Based on experimental investigations 193 colorectal carcinoma patients (UICC stages I-III) were enrolled in a prospectively randomized clinical trial. 93 patients were treated perioperatively with D-galactose- (treatment group: 1.5 g/kg body weight and per day) or D-glucose containing electrolyte infusions (control group: n = 100). Significant side effects were not observed. There were no cases of perioperative mortality. The overall complication rate was 7.3%. Since tumor stages were unequally distributed, analysis was performed in strata. Patients were observed for a total of 6237 months. Differences in overall survival and survival free of recurrence and hepatic metastases were negligible for stages I and II. For stage III carcinoma patients (n = 75) analysis of survival free of hepatic metastases revealed a shift to delayed events (i.e. hepatic metastases or death) after D-galactose treatment within 24 months following surgery. In patients with stage III carcinoma there was an indication for an overall benefit in survival after D-galactose treatment (p = 0.102).