Rollover accidents are among the most difficult of the major accident categories to analyze through the application of full-scale test procedures. Because of this difficulty, industry and governments worldwide have not reached a consensus on a full-scale test technology to assess a vehicle’s ability to resist rollover motion. Additionally, progress in developing test procedures to define rollover resistance has been retarded by the factors of accident complexity and a reluctance to separate maneuver handling issues from those specifically related to the rollover event. The relevance of field accident data to test procedure selection will be described with examples from US. experience. Candidate procedures from around the world will be discussed and a set of criteria for judging and comparing the merits of these proposals will be suggested. All procedures represent compromises between the desire to include every contributing factor and a need to distinguish differences in rollover performance. Also, problems associated with validating candidate procedures with field accident data are described.