The intent of crashworthiness standards and new car evaluations based on full vehicle testing is to promote designs that eliminate many, if not all, of the serious injuries in real-world crashes of equivalent or lesser severity to that used in the test. A sample of real-world frontal crashes in the United States was examined and the estimated crash severities were related to impact speeds for single-vehicle offset crashes into deformable barriers. For each of the vehicles in these tests, the postcrash vehicle deformation was measured using the procedures specified for CRASH3 and a delta V was calculated. The results of this study suggest that a 40 percent offset test into a deformable barrier at 64 km/h represents a real-world crash severity below which about 75 percent of all MAIS 3 or greater injuries and slightly less than half of all fatal injuries to passenger car occupants in frontal offset crashes occur in the United States. The fact that many deaths and serious injuries occur in higher severity crashes suggests that this test speed choice is not too high and that standards or crashworthiness evaluations in offset tests into deformable barriers at significantly lower speeds would be ignoring large numbers of real-world crashes with serious injuries.