Two prospective randomized trauma trials have shown recombinant factor VIIa (rFVIIa) to be safe and to decrease transfusion requirements. rFVIIa is presently used in 22% of massively transfused civilian trauma patients. The US Military has used rFVIIa in combat trauma patients for five years, and two small studies of massively transfused patients described an association with improved outcomes. This study was undertaken to assess how deployed physicians are using rFVIIa and its impact on casualty outcomes. US combat casualties (n = 2,050) receiving any blood transfusion from 2003 to 2009 were reviewed to compare patients receiving rFVIIa (n = 506) with those who did not (n = 1,544). Propensity-score matching (primary analysis) and multivariable logistic regression were used to compare outcomes. Differences were determined at p < 0.05. Twenty-five percent of patients received rFVIIa. Significant differences were noted between groups in indices of injury severity (Injury Severity Score, Abbreviated Injury Scale score, and Glasgow Coma Scale score), admission physiology (systolic blood pressure, diastolic blood pressure, heart rate, temperature, base deficit, hemoglobin, and international normalization ratio), and use of blood products, indicating that patients treated with rFVIIa were more severely injured, in shock, and coagulopathic. For propensity-score matching, factors associated with death were used: Injury Severity Score, Glasgow Coma Scale score, heart rate, systolic blood pressure, diastolic blood pressure, Hgb, and total packed red blood cell. A total of 266 patients per group were matched; 52% of the rFVIIa group. After pairing, there were no significant differences in any of the demographics, including incidence of massive transfusion (53% vs. 51%). There was no difference in the rate of complications (21% vs. 21%) or mortality (14% vs. 20%) for patients not treated or receiving rFVIIa, respectively. In military casualties, rFVIIa is used in the most severely injured patients based on physician selection rather than on guideline criteria. Use of rFVIIa is not associated with an improvement in survival or an increase in complications. The undetected bias of physician selection of patients for treatment with rFVIIa, likely, has an impact on case matching to achieve equivalence similar to that of randomized control studies. This inability to match populations, thus, prevents definitive interpretation of this study and others studies of similar design. This problem emphasizes the need to develop entry criteria to identify patients who could potentially benefit from use of rFVIIa and the need to subsequently perform efficacy studies.