Increasing resolution of neuroimaging techniques has allowed investigators to track changes in neural activity not only resulting from, but also in anticipation of choice. This has supported identification of circuits that respond to the magnitude and probability of expected gains and losses, as well as determination of whether their activity can also predict upcoming choice. Even more recent evidence suggests that beyond predicting individual choice, activity in some of these circuits may forecast aggregate choice. Although not a perfect match, these collected findings imply that in less than two decades, researchers are converging on a neural basis for expected utility.