Cerebral autoregulation is the body’s ability to control the flood of blood into the brain in response to changing blood pressure. Monitoring of this system has been linked with post-incident prognosis in patients for a number of medical conditions such as stroke and traumatic brain injury. This monitoring can be achieved through a number of different popular analysis techniques such as transfer function analysis, the autoregulation index and the correlation coefficient. Metrics of autoregulation from all these methods exhibit variability, due both to errors in the method and also potential physiological causes such as fluctuations in the response over time. This project focuses on first developing techniques that aim to improve monitoring of autoregulation by providing a means of assessing the estimate variability, and secondly developing our understanding of some of the characteristics of autoregulation such as these fluctuations over time. A new method that uses a parametric-bootstrap approach to generate confidence intervals of autoregulation estimates, allowing for determination of the robustness of said estimates, has been developed, with examples of its potential applications being outlined. These applications include assessing the impact of poor-quality data on the variability of autoregulation estimates, and the optimisation of assessment parameters to obtain the most robust results. This method was then used in conjunction with another developed method to determine the presence of non-stationarity in autoregulation. By generating distributions using the bootstrap approach and using a newly developed resampling approach to calculate whether distributions are significantly different from each other, the level of non-stationarity present in a cohort of recordings can be assessed, providing meaningful information that is important to the clinical applications of autoregulation monitoring. Finally, the difference between how autoregulation responds to small and large changes in pressure has been examined, in order to determine whether or not monitoring of the autoregulation of vulnerable patients without the inducement of any extreme pressure change is suitable in determining how their bodies will respond to sudden pressure swings.