Volunteer subject studies in low-speed rear impacts have shown that significant lumbar spine injuries are unlikely in such collisions. Anthropomorphic test devices (ATD) used in low to medium speed rear impact simulations have similarly revealed an unlikely mechanism to cause lumbar spine injuries. However, low back complaints after rear impacts are common in clinical practice. We attempt here to determine the incidence of lumbar spine injuries from actual field data which may provide an insight into the apparent paradox between experimental data and clinical practice.We examined the incidence of all spine injuries in the NASSCDS (National Automotive Sampling System - Crashworthiness Data System) database from 1993 to 2009. We limited the data to only look at rear-end crashes involving two vehicles. We analyzed crash severity (delta-V), occupant injuries by AIS (Abbreviated Injury Scale) code, seat performance and restraint use in over 7,500 (with a weighted value of 2.5 million) passenger vehicle accidents. Of the 7,500 accidents, approximately 500 (weighted value of 225,000) passengers reported some type of lumbar spine injury. These injuries included strains or sprains to the spine (AIS 1 injuries), fractures and herniations (AIS 2). In particular, we stratify the incidence of lumbar spine injuries to low, medium and high speed rear impact crash severities, and correlate seatback deformation and restraint use to the lumbar injuries.The analysis indicated the small number of reported lumbar injuries associated with rear end collisions were mostly musculoskeletal strains/sprains and a few fractures and herniations. Furthermore, there was no correlation of higher incidence of lumbar injury with increasing delta-V. The results of the analysis agreed with those from low-speed human subject tests and low- to moderate-speed ATD rear end simulations that concluded the lumbar spine is well protected by the seat back for properly seat-belted passengers.