The development of a repeatable dynamic rollover test methodology with meaningful occupant protection performance objectives has been a longstanding and unmet challenge. Numerous studies have identified the random and chaotic nature of rollover crashes, and the difficulty associated with simulating these events in a laboratory setting. Previous work addressed vehicle level testing attempting to simulate an entire rollover event but it was determined that this test methodology could not be used for development of occupant protection restraint performance objectives due to the unpredictable behavior of the vehicle during the entire rollover event. More recent efforts have focused on subsystem tests that simulate distinct phases of a rollover event, up to and including the first roof-to-ground impact. This paper explores two subsystem rollover test methods up to the first roof-to-ground impact, and highlights the importance of simulating the lateral vehicle acceleration that occurs during the pre-trip/trip phase of a rollover event in order to provide more realistic trailing-side Anthropomorphic Test Device (ATD) kinematics and enable more meaningful restraint evaluations.