Twelve full-scale tests have been performed to investigate the underride problem and to determine the effectiveness of several specific guard designs in preventing underride type collisions. Initial tests were performed using a rigid flat plate underride guard positioned against an SAE barrier. Later tests were conducted using specific guard designs mounted on the rear of tandem axle flatbed and van trailers. Parameters studied include impacting vehicle size and weight, impact velocity, underride guard ground clearance height, rigid and yielding guards, frame and frameless (monocoque) trailer target vehicles, and symmetrical (on-center) and unsymmetrical (off-center) impacts. Tests were conducted at impact velocities from 30-40 mph. Guard heights of 18 and 24 in. were investigated using cars ranging in weight from 1600-5150 lb. Data obtained from these tests include deceleration of the impacting automobiles, loading experienced by the underride guards, and high-speed motion pictures of the collision sequences.In addition to the experimental effort, mathematical modeling of the underride collision was performed using a system of discrete masses and springs to represent the impacting car and truck underride guard. Crash test data were used to define the force-deflection properties for the various springs representing structural components of the automobile. Simulations were run for a series of underride guard force magnitudes and wave forms to examine the effects of underride guard force characteristics on the impacting vehicle.Conclusions, based on the results of the full-scale crash tests and the computer simulations, are drawn concerning various aspects of the underride problem such as guard strength requirements, guard ground clearances, and the force-stroke requirements for yielding guard designs. Effects of differences in car weights and sizes, and truck-trailer structures, on the general underride guard problem are also considered.