Objectives: To 1) characterize pre-cesarean blood bank testing, 2) describe the transfusion experience in a large series of cesarean patients, and 3) evaluate safety and cost implications of a ''hold clot'' order for patients at low risk for transfusion. Methods: A review of 1111 consecutive cesarean patients used computerized perinatal and blood bank data bases and a detailed chart review of all cross-matched patients. Information collected included indications for cesarean and transfusion, etiology of hemorrhage, transfusion number and type, admission and lowest hemoglobin level, and information regarding the events leading to transfusion. A blinded review of the cross-matched patient's information assessed whether a cross-match was appropriate or could have been replaced safely by a ''hold clot'' (current clot tube in blood bank) order. Results: Nineteen patients (1.7%) were transfused. The only patients requiring a transfusion were diagnosed with placenta previa, placenta accreta, anemia, preeclampsia/hemolysis, elevated liver enzymes, low platelets (HELLP syndrome), or hemorrhage. A comparison of two blood banking approaches (routine pre-cesarean type and screen testing versus a ''hold clot'' order for cesarean patients at low risk for transfusion) indicated that the latter would reduce costs by $45 per cesarean, or $95,000 annually. Conclusions: The incidence of transfusion was low (1.7%) and associated with specific diagnoses (previa, accreta, anemia, preeclampsia/HELLP, or hemorrhage). The data support the replacement of pre-cesarean type and screen testing with a ''hold clot'' order for patients at low risk for transfusion with negative prenatal antibody screen. This approach is safe and would reduce cost substantially.