This paper describes an analysis of roof contacts in inverted impacts. The proven methodology of lumped mass modeling is used to evaluate the effect of a number of variables on neck loading. Several parameters were evaluated including roof strength, drop height, head clearance, restraint use and stiffness, roof padding and neck stiffness.In general, the simulation results agreed well with existing head first impact test data. However, interesting results were obtained when neck stiffness was varied to represent the difference between human-like and Hybrid III spring rates. Biomechanical data generally agrees, independent of test methodology, that the Hybrid III neck is at least 10 times as stiff as a cadaver neck. Using an approximation of the Hybrid III neck stiffness (5.25 KN/cm) generally resulted in initial neck loads exceeding “upper” human tolerance level of 6500 N. Using a neck stiffness of 10 percent of the Hybrid III resulted in neck loads below the “lower” human tolerance level of 4,000 N.Hybrid III force levels were more or less the same independent of roof stiffness. The lower stiffness neck force levels were similar in magnitude; although, longer in duration, with a relatively weak roof. Since neck injury potential is, apparently, a function of both magnitude and duration, this modeling indicates the potential for a weak roof to increase the likelihood of injury in a rollover collision.