The effect of perioperative blood transfusions on survival after curative resection for colorectal carcinoma was studied retrospectively on 270 patients with special emphasis on whole blood transfusion. Eighty-six (32 percent) patients did not receive blood transfusions, whereas the remaining 110 (41 percent) received packed cells and 74 (27 percent) received at least 2 units of whole blood. The recurrence-free and overall 5-year survival rates for the transfused group were 58 and 57 percent, respectively, and for the non-transfused group, 78 and 72 percent, respectively. These differences were highly significant (log rank test, P less than 0.001). The recurrence-free and overall 5-year survival rates for patients who received only whole blood or a combination of whole blood and packed cells was 51 percent, and of those who received only packed cells was 63 percent. The difference was not statistically significant (log rank test, P = 0.08, P = 0.35). Administration of more than 6 units of blood is attended with a significantly worse 5-year disease-free survival rate than administration of a smaller volume. The survival was not decreased if 6 or more units of whole blood were given. The authors conclude that perioperative blood transfusions have a significant detrimental effect on survival after curative resection for colorectal carcinoma. This effect is not enhanced by whole blood transfusions.