Although structurally and functionally constrained, the human brain displays a high level of plasticity, which is most prominent during the first months and years of life, but persists throughout the lifespan. This is paired with our protracted ability to learn novel skills and acquire novel information. While most studies on brain plasticity focus on the acquisition of extremely simple sensory or motor skills, very little is known about the local and global changes that the brain undergoes while learning more complex, symbolic, culturally-mediated skills, such as reading or arithmetical calculation. There are two different ways in which we can learn to solve arithmetical calculations: via rote memorization of specific operand-result associations (e.g., 4 x 5 = 20), or via the automated use of strategies that emerge from deep understanding of arithmetic principles and number semantics (e.g., 4 x 5 = [4 x 10] / 2 = 20). Although these different learning regimes might result in equally effective learning (at least in the normal population), the neural circuitries that are involved and possibly show learning-related changes are likely to be substantially different. Our aim with this project is to investigate whether and how learning to solve novel arithmetical operations entail changes in the brain and whether different training regimes (strategy-based vs. memory-based) entail specific neuro-functional brain reorganizations. Gathering insights into the brain changes underlying arithmetic learning will be fundamental not only to understand normal functioning but also to interpret changes in normal aging and patterns of disruption in brain-damaged patients, and it may also help highlighting ways for improving teaching and rehabilitation programmes.