Repetitive traumatic brain injury (TBI) occurs in a significant portion of trauma patients, especially in specific populations, such as child abuse victims or athletes involved in contact sports (e.g. boxing, football, hockey, and soccer). A continually emerging hypothesis is that repeated mild injuries may cause cumulative damage to the brain, resulting in long-term cognitive dysfunction. The growing attention to this hypothesis is reflected in several recent experimental studies of repeated mild TBI in vivo. These reports generally demonstrate cellular and cognitive dysfunction after repetitive injury using rodent TBI models. In some cases, data suggests that the effects of a second mild TBI may be synergistic, rather than additive. In addition, some studies have found increases in cellular markers associated with Alzheimer's disease after repeated mild injuries, which demonstrates a direct experimental link between repetitive TBI and neurodegenerative disease. To complement the findings from humans and in vivo experimentation, my laboratory group has investigated the effects of repeated trauma in cultured brain cells using a model of stretch-induced mechanical injury in vitro. In these studies, hippocampal cells exhibited cumulative damage when mild stretch injuries were repeated at either 1-h or 24-h intervals. Interestingly, the extent of damage to the cells was dependent on the time between repeated injuries. Also, a very low level of stretch, which produced no cell damage on its own, induced cell damage when it was repeated several times at a short interval (every 2 min). Although direct comparisons to the clinical situation are difficult, these types of repetitive, low-level, mechanical stresses may be similar to the insults received by certain athletes, such as boxers, or hockey and soccer players. This type of in vitro model could provide a reliable system in which to study the mechanisms underlying cellular dysfunction following repeated injuries. As this area of TBI research continues to evolve, it will be imperative that models of repetitive injury replicate injuries in humans as closely as possible. For example, it will be important to model appropriately concussive episodes versus even lower level injuries (such as those that might occur during boxing matches). Suitable inter-injury intervals will also be important parameters to incorporate into models. Additionally, it will be crucial to design and utilize proper controls, which can be more challenging than experimental approaches to single mild TBI. It will also be essential to combine, and compare, data derived from in vitro experiments with those conducted with animals in vivo. These issues, as well as a summary of findings from repeated TBI research, are discussed in this review.