Neither the experimental nor analytical techniques currently used to study vehicle rollover accidents accurately represent most actual rollovers. Until recently, crash tests to study rollovers have used either snubbed dollies or guided ramps to cause rollovers. Real world vehicle rollovers, however, are caused by a variety of mechanisms, including impacting curbs or obstacles, sliding through soil or sod, and dropping off embankments. Analytical methods proposed to model rollover events represent idealized curb trip situations and provide unrealistically low estimates for the lateral speeds needed to cause rollovers.This paper presents the results of a continuing investigation into the mechanics of real world vehicle rollovers. Rollover tests with vehicles tripped by a curb, sliding in soil, and thrown from a dolly are presented. The mechanics of the different trip modes is discussed and a simple analytical model to represent the trip mode behavior presented.