Abstract: A critical feature of adaptive behavior is the ability to seek rewarding experiences and avoid punishing ones. This ability, however, is markedly challenged by both external and internal factors. For example, an experience that we associate with rewarding outcomes, like going to your favorite restaurant, can be variable. This could be part of the expected fluctuation of the food served at the restaurant (i.e. most of the time the meals are enjoyable, but sometimes they are less so), or it could be indicative of an underlying change in the restaurant’s quality (e.g. the chef has changed). How we interpret outlying experiences (known to cause “reward prediction errors” in reinforcement learning), and thus learn and remember the outcomes of these experiences, determines a host of daily and long-term decisions. In this talk I will present three series of behavioral experiments that characterize the effects of reward prediction errors on the structure of memories, and will also present data showing how internal factors, such as depressive symptoms, can modulate these effects.