Cerebral autoregulation describes the process by which cerebral blood flow is maintained despite fluctuations in cerebral perfusion pressure. The assessment of cerebral autoregulation is a key to the optimisation of cerebral perfusion pressure in patients with brain injury. This review evaluates the current evidence for transcranial Doppler in the assessment of cerebral autoregulation. The study of cerebral autoregulation classically assesses changes in cerebral perfusion pressure secondary to changes in systemic blood pressure. It is defined static autoregulation if blood pressure changes are progressive, thereby allowing a steady-state autoregulatory response to be completed. For sudden changes in blood pressure, the autoregulatory response is defined as dynamic. The static and dynamic components of cerebral autoregulation have been approached using linear mathematical models (models based in direct correlations). Over the past decade, demonstration of the nonstationary (the property of changing over time or space) behaviour of cerebral autoregulation has emphasised the benefit obtained in using nonlinear statistical models (models based on changeable functions), suggesting that these methods may improve the mathematical representation of cerebral autoregulation. Despite the multiple determinants involved in cerebral autoregulation, it appears feasible to reliably assess cerebral autoregulation through the combination of linear and nonlinear methods. Nonlinear methods appear attractive in the research setting, but the challenge is how to adopt these methods to the clinical setting.