Recent research in social psychology suggests that our attitudes and actions don’t typically issue from our reflective capacities. Instead, it appears that they often issue from what psychologists refer to as “automatic” processes, understood as psychological processes that operate quickly, efficiently, and outside of our conscious awareness and control. Much of this same research also suggests that when we do exercises our reflective capacities (rare as it may be), it’s usually not to question our attitudes and actions but rather to rationalize them (that is, to come up with reasons to do, think, and feel things that we want to do, think, and feel anyway). In light of these findings, a number of people have recently argued that we aren’t the kind of rational, reflective creatures that philosophers have traditionally thought we are. My aim in this dissertation is to respond to their worries. In the first half, I focus on questions related to our ability to do things for reasons. Common sense suggest that we often do things for reasons, but if our attitudes and actions typically issue from automatic processes, to what extent is this true? I argue that the best way to make good on the idea that we often do things for reasons is to accept both that our ability to do things for reasons doesn’t depend on our reflective capacities and that many of the processes that actually guide our attitudes and actions, though perhaps “automatic,” are nevertheless quite responsive to reasons. In the second half of the dissertation, I focus on whether we can rely on our reflective capacities to improve our agency. Although I deny that we need these capacities to do things for reasons, the reasons for which we do things aren’t always good reasons, and it’s only natural to think that we can rely on our reflective capacities to monitor and control their influences on us. However, this sort of view faces to main empirical challenges. First, it appears that we often don’t know what the causes of our attitudes and actions are, so how can we be expected to monitor and control their influences on us? Second, even if we do know what the causes of our attitudes and actions are, why should we think that we’re inclined to use reflection to question them? Isn’t it much more likely that we’ll use it to rationalize them instead? In response to the first challenge, I argue that once we recognize that reflection doesn’t require us to have introspective access to its objects, it becomes more plausible to think that we’re often in a good position to reflect on the causes of our attitudes and actions. In response to the second challenge, I argue that as long as we focus on the ways in which other people improve our reflection, the idea that we’ll often be inclined to use reflection to question our attitudes and actions becomes more plausible as well.