In the roadside safety community, there has been debate over the influence of vehicle and barrier type on rollover rates in traffic barrier crashes. This study investigated rollover rates between sport utility vehicles (SUVs), pickup trucks, and cars in vehicle-traffic barrier crashes and has examined the effect of barrier type on rollover risk for concrete barrier and metal barrier impacts. The analysis included 955 barrier impact cases that were selected from 11-years of in-depth crash data available through the National Automotive Sampling System (NASS) / Crashworthiness Data System (CDS). In real world tow-away level longitudinal barrier collisions, the most important predictors of vehicle rollover were found to be vehicle type and whether the vehicle was tracking prior to barrier impact. Based on binary logistic regression, SUVs were found to have 8 times the risk of rollover as cars in barrier impacts. Although pickups were found to have an increased risk of rollover compared to cars, the risk was not as pronounced as that found for SUVs. This finding has direct implications for the full scale crash testing of longitudinal barriers as the testing procedures have been predicated on the assumption that the pickup truck provides a critical or worst case impact scenario. In towaway crashes, our study does not support the notion that concrete barriers have a higher risk of vehicle rollover than metal beam barriers.