With clinical datasets continuing to grow in size and complexity, there has been an increased interest in using powerful computational tools such as machine learning (ML) for the analysis of complicated associations to provide an important insight to the clinician. Recent neurosurgical literature has exemplified this interest, as there has been a significant rise in the breadth and depth of ML applications to a host of areas in neurosurgery. These include neuro-oncology, stroke, epilepsy, neurotrauma, and spine surgery. Within these areas, ML models have been developed to aid in clinical settings encompassing the entire operative timeline, including preoperative neurosurgical diagnostics, intraoperative management, and postoperative prediction of outcomes. The recent adoption of ML in neurosurgery has occurred due to ML’s numerous advantages over conventional statistical tools. These include ML’s ability to handle large datasets, decipher important predictors without a priori specification, and determine complex nonlinear relationships within data. However, the promise behind ML is not without limitations, as ML algorithms are heavily reliant on high-quality data and carry important medico-legal ramifications if used without regulation. This chapter will discuss the current applications of ML to neurosurgery across multiple areas, potential future directions that ML could take, and the important limitations of these computational tools. Via this discussion, we hope to provide the reader with an insight into the potentially significant impact that ML can have on neurosurgery.