Animal research has indicated an important role for the ventral tegmental area (VTA) and its neurotransmitter dopamine in reinforcement learning, an elementary form of learning on the basis of rewards and punishments. In recent years, cognitive neuroscientists have used these animal data to develop various highly sophisticated theories on the role of the dopamine system in human reinforcement learning. However, crucial empirical tests of these theories in human subjects have generally been lacking, probably in part because of the considerable methodological challenges involved. The goal of the proposed research program is to address these challenges using a number of novel methods: (i) High-field (7 Tesla) functional magnetic resonance imaging in combination with a set of recently published acquisition and analysis methods will allow us to directly measure activity of the VTA; and (ii) pharmacological magnetic resonance imaging will allow us to investigate the effects of experimental manipulations of the dopamine system on brain function and behavior. We will use these methods to test detailed theoretical predictions regarding the role of the human dopamine system in coding error responses and negative feedback, in learning from positive and negative outcomes, and in valuing immediate versus delayed rewards. The proposed research will be important in elucidating the role of the dopamine system in human reinforcement learning. At the methodological level, it will open the door for the careful study of the role of dopamine in human cognition at large.