Traumatic brain injury (TBI) continues to be an important cause of mortality and morbidity, but its pathophysiology is no longer considered an instantaneous irreversible event occurring at the time of injury. Therein, neuroprotection is the attempt to salvage sublethally injured neurons which subsequently die in post-primary sequelae. Key to the discovery of neuroprotective strategies is the development of reliable models of brain injury--both in vivo and in vitro. While numerous studies on in vivo animal models have yielded encouraging results, these have largely failed to translate effectively in humans. One approach out of this impasse may be to re-explore in vitro models to dissect out specific pathophysiological mechanisms and only then test clearer hypotheses on in vivo models, which are more likely to subsequently translate into neuroprotective therapies of the future. Moreover, milder forms of TBI are a more realistic target for therapeutic intervention as more is understood about the vulnerability of surviving neurons and the capacity to salvage them. Several types of injury models are described including transection, compression, barotrauma, acceleration, hydrodynamic and cell stretch models with their advantages and disadvantages discussed in turn, as well as a survey of the cell cultures used, namely immortalized cell lines, primary cultures and organotypic (explant) cultures. We emphasize advances in three-dimensional strain simulation and a recent interest in modelling milder injuries, and argue that in vitro models may be a useful complement to in vivo models in studying TBI.