The transformative effects of practice on performance is acknowledged across domains. Several theories of skill acquisition (e.g., Anderson, 1982; Logan, 1988; Rickard, 1997) have offered explanations for how practice reduces speed and increases accuracy of task performance. These theories, however, differ by the degree in which speedup is attributed to either changes in or increased efficiency of the cognitive processes executed to solve a problem. Attempts to evaluate the different models of skill acquisition have often been limited by difficulties in identifying strategy shifts in the transition from computation to retrieval. By using a multi-pronged approach that applies brain imaging, cognitive modeling, and machine learning methods, Tenison examines three models of skill acquisition and explore in depth the cognitive changes that occur with practice. In the first study, Tenison uses an unsupervised modeling method to examine the different possibilities in the learning function fit, as well as, evaluate the number of learning phases according to how well they fit the data at the individual item level. In the second study, Tenison uses neuroimaging data (i.e., functional magnetic resonance imaging) in combination with machine learning techniques to explore the distinction between the cognitive processes involved in the phases of skill acquisition and study how practice impacts cognitive processing within a learning phase. In study 3, Tenison uses magnetoencephalography data to explore the degree to which learning alters specific cognitive processes and the speed at which those processes are used. Collectively, this set of experiments helps develop a more nuanced understanding of how practice impacts learning and has implications for in how we can support skill acquisition.