On December 3, 2003, a single-car impact test was conducted to assess the crashworthiness performance of a modified passenger rail car. A coach car retrofitted with a crash energy management (CEM) end structure impacted a fixed barrier at approximately 35 mph. This speed is just beyond the capabilities of current equipment to protect the occupants. The test vehicle was instrumented with accelerometers, string potentiometers, and strain gages to measure the gross motions of the car body in three dimensions, the deformation of specific structural components, and the force/crush characteristic of the impacted end of the vehicle. The CEM crush zone is characterized by three structural components: a pushback coupler, a sliding sill (triggering the primary energy absorbers), and roof absorbers. These structural mechanisms guide the impact load and consequent crush through the end structure in a prescribed sequence. Pre-test activities included quasi-static and dynamic component testing, development of finite element and collision dynamics models and quasi-static strength tests of the end frame. These tests helped verify the predicted structural deformation of each component, estimate a force-crush curve for the crush zone, predict the gross motions of the car body, and determine instrumentation and test conditions for the impact test. During the test, the passenger car sustained approximately three feet of crush. In contrast to the test of the conventional passenger equipment, the crush imparted on the CEM vehicle did not intrude into the passenger compartment. However, as anticipated the car experienced higher accelerations than the conventional passenger car. Overall, the test results for the gross motions of the car are in close agreement. The measurements made from both tests show that the CEM design has improved crashworthiness performance over the conventional design. A two-car test is performed to study the coupled interaction of CEM vehicles as well as the occupant environment. The train-to-train test results are expected to show that the crush is passed sequentially down the interfaces of the cars, consequently preserving occupant volume.