Autologous transfusion is an option infrequently used in the UK, partly because the direct costs are greater than for allogeneic transfusion. An analysis of the cost consequences of substituting autologous for allogeneic blood, by using cell salvage, was undertaken from a hospital perspective. Direct costs were estimated for two different cell salvage devices and sensitivity analysis performed on the key variables. Allogeneic transfusion may be associated with increased rates of postoperative infection due immunomodulation and immunosuppression. As a result, one of the short-term benefits of autologous transfusion is a possible reduction in length of hospital stay. This was the most important variable affecting the cost of autologous transfusion. Cost equivalence for autologous and allogeneic blood was reached at reductions in hospital stay of between 0.3 and 2 days across a range of variables. Reductions in length of hospital stay of this magnitude have been reported in several small studies. The results of our cost analysis suggest that autologous transfusion should not be rejected on the grounds of cost and we recommend a large-scale randomized controlled trial of autologous vs. allogeneic transfusion.