The research study consisted of Phase II of the development and full-scale vehicle crash testing of a bullnose barrier concept for the treatment of median hazards. The bullnose guardrail consisted of a 12-gauge thrie beam rail supported by twenty-two wood posts, eleven posts on each side of the system. Horizontal slots were cut in the valleys of selected thrie beam sections to aid in vehicle capture as well as to reduce the buckling and bending capacities of the rail. Two full-scale crash tests were performed, both using a 2000-kg pickup truck. The first crash test, impacting at a speed of 100.2 km/h and an angle of 0 degrees, was unsuccessful following the rupture of the thrie beam and subsequent uncontrolled penetration of the vehicle behind the barrier. Computer simulation of the failed test was performed using LS-DYNA. Analysis of the simulation and full-scale test results led to the addition of two steel cables placed behind the top and middle humps of the thrie beam nose section to aid in truck containment without stiffening the barrier. Computer simulation of the modified design demonstrated successful containment of the pickup truck. The second test, impacting at a speed of 103.5 km/h and an angle of 0 degrees was determined to be successful according to the safety standards set forth by the Test Level 3 (TL-3) evaluation criteria described in the NCHRP Report No. 350, "Recommended Procedures for the Safety Performance Evaluation of Highway Features" (1993). The next phase of the bullnose barrier system design will be to complete the remaining crash tests needed for compliance with the NCHRP Report No. 350 safety standards for median barriers.