Background: Improvised tourniquets may be used to treat limb wound hemorrhage, but there is little evidence for best techniques of use. The purpose of the present study is to compare use of two techniques of improvised tourniquet application and use of a common commercial tourniquet that is nonimprovised. Methods: A laboratory experiment was conducted to assess three groups of strap-and-windlass tourniquet designs on a manikin to test for differences in performance. Groups included two types of improvised tourniquets (bandage and bandana) and a third group that served as a control, the commercial Combat Application Tourniquet. Two users performed 10 tests of each group. Results: The commercial CAT had 100% effectiveness, but both improvised tourniquets had poor effectiveness (40% and 10% for the bandage and bandana groups, respectively). The commercial CAT performed fastest; the two improvised tourniquet groups were slower than the commercial group (p < .0001, both) but were not statistically different from each other. All time-of-application results in the commercial group were less than the minimums of either improvised group. The commercial CAT had the highest mean pressures, and all such pressures were within safe and effective ranges. Low pressures generated by both improvised tourniquet groups were ineffective. All results of simulated blood loss with the commercial CAT group were less than the minimums of either improvised tourniquet group. Conclusion: In the present experiment, the commercial CAT performed better than either improvised tourniquet.