During the period of 1978 to 1986, 66 patients (31 men, 35 women) with a mean age of 28.4 years and various sickle cell hemoglobinopathies underwent 82 surgical procedures; 28 were emergencies. Fifty of the 66 patients had HbSS, 13/66 had HbSC, and 3/66 had HbS-thalassemia. All 66 patients received transfusions, although not for all procedures. In 48 patients, transfusion therapy was only administered preoperatively. Simple transfusions (1 to 10 units) were administered in 31 of 48 procedures. Exchange transfusions (1 to 6 units) were performed in nine of 48 procedures. Preoperative hematocrit ranged from 7.0% to 54.2%; of those receiving transfusions the hematocrit ranged from 22.6% to 53.7%. Intraoperative transfusions (1 to 10 units) were performed in 14 of 82 procedures; postoperative transfusions (1 to 6 units) were performed in 13 of 82 procedures. No advantage was noted in preoperative exchange transfusion as measured by a decrease in postoperative complications; a slight increase was seen in atelectasis in this group of patients with preoperative transfusions. An increase was reported in the complication rate of patients with an hematocrit of less than 30%. The type of transfusion (preoperative, intraoperative, or postoperative) administered did not appear to be related to postoperative morbidity rates. The complication rate for simple transfusions was 51.6% and for multiple transfusions, 55.6%. HbSS hemoglobinopathy had the higher complication rate. The hepatitis B surface antigen was demonstrated in four of 66 (6.1%) patients; ten of 66 (15.2%) developed alloantibodies. The benefits of transfusion therapy should be judged according to clinical needs; not all sickle cell patients need exchange or preoperative transfusion.