The choices we make are central to who we are, not only because their outcomes shape us but also because they are informed by our pasts. Indeed, rather than taking in objective information and translating it into action, human decision makers draw on previous experiences to form subjective predictions about future outcomes. These predictions serve as the foundation for evaluating options and ultimately drive choice. In this dissertation, I highlight three different ways in which past experience, accessible to us in a variety of forms, enables value-based decision-making: through inferences about the temporal dynamics of reward delivery (Chapter 2), through trial-and-error learning of the value of choice options (Chapter 3), and through the encoding and retrieval of details from episodic memory (Chapter 4). In Chapter 1, before delving into the empirical data, I provide an overview of the current neuroscientific evidence in favor of specialized brain regions that support our ability to encode and retrieve information, update our expectations, and use these expectations to guide valuation. In Chapter 2, I focus on data collected in a sample of participants with lesions to different areas to the prefrontal cortex. Using behavioral data and computational modeling, I show that lesions to the vmPFC cause people to wait less for delayed rewards, while lesions to the dmPFC and anterior insula lead to impairments in the adaptive calibration of persistence, and lesions to lateral portions of the PFC do not alter waiting behavior relative to healthy controls. In Chapters 3 and 4, I turn towards fMRI data collected in a sample of younger and older adults while they completed decision tasks that were designed to recruit different forms of memory. Chapter 3 focuses on incremental learning of the value of choice options, as in the case of reinforcement learning. I find that despite the procedural nature of this task, precise stimulus representations of choice options in the hippocampus predicts accurate value-based choice. Furthermore, older adults perform less well than younger adults – a performance decrease I trace back to reduced connectivity between the hippocampus and value-responsive parts of the brain. In Chapter 4, I turn towards social decisions from episodic memory and show that older adults are less able to retrieve unique past experiences in order to make decisions about whom to interact with. Instead, I identify neural correlates of an increased reliance on irrelevant facial features like how generous a person looks. Taken together, these studies emphasize the interdependencies between learning, memory, and decision-making by revealing how disruptions to specific neural circuits alter how we evaluate and select choice options. In Chapter 5, I conclude by highlighting some of the challenges with this modular approach to understanding the brain and consider what a more parsimonious view of understanding neural functions may look like.