Background. Excess ordering of blood products for surgical cases is expensive and wasteful. Evidence has shown that institution-specific versions of the Maximum Surgical Blood Order Schedule (MSBOS) lead to better ordering practices. Most MSBOSs recommend a crossmatch for a minimum of 2 units of packed red blood cells (PRBCs) for cardiac surgical cases; however, studies have shown that >50% of these patients receive no transfusions. Our aim was to create a blood order algorithm for cardiac surgical cases that would decrease unnecessary crossmatching. Methods. Retrospective data was collected for 264 patients from January 2011 through April 2012. The crossmatch-to-transfusion ratio (C:tx), transfusion probability (%T), and transfusion index (TI) were calculated for each type of procedure. Results. All 264 patients were crossmatched and 98 patients were transfused, resulting in an overall transfusion probability (%T) of 37.12% (95% confidence interval 31.52-43.09). A total of 1175 units of blood were crossmatched, but only 370 units of blood were transfused, resulting in a C: tx of 3.17 (95% confidence interval 2.61-4.03). The average number of units transfused per procedure (transfusion index) was 1.40. C: tx was highest and TI was lowest for CABG, where approximately 11 units of blood were ordered for every 1 unit transfused (C: tx = 11.70 +/- 3.04), and the TI was 0.32. Conclusions. Using the gold standard C: tx of >2:1 as an indicator of inappropriate blood utilization, our analysis confirmed that excessive crossmatching occurred for several procedures. Now a subset of cardiac surgical cases only requires a type and screen order prior to surgery.