We developed a quantitative method to study the efficacy of intravenously delivered human factor VIII in hemophilia A mice. Mortality was assessed after tail transection under conditions in which there were no survivors in untreated hemophilia A mice. Blood loss was significantly greater in untreated hemophilia A mice compared to normal C57BL/6 mice, and in hemophilia A non-survivors that were treated with sub-therapeutic doses of factor VIII. The up-and-down method for small samples yielded an estimated dose of factor VIII producing survival in 50% of the mice (ED(50)) of 58 units/kg (95% confidence interval: 42.4-78.5 units/kg). This method should be useful in the evaluation and comparison of novel factor VIII products delivered either parenterally or in a gene therapy setting.