Learning what to approach, and what to avoid, involves assigning value to environmental cues that predict positive and negative events. Studies in animals indicate that the lateral habenula encodes the previously learned negative motivational value of stimuli. However, involvement of the habenula in dynamic trialby-trial aversive learning has not been assessed, and the functional role of this structure in humans remains poorly characterized, in part, due to its small size. Using high-resolution functional neuroimaging and computational modeling of reinforcement learning, we demonstrate positive habenula responses to the dynamically changing values of cues signaling painful electric shocks, which predict behavioral suppression of responses to those cues across individuals. By contrast, negative habenula responses to monetary reward cue values predict behavioral invigoration. Our findings show that the habenula plays a key role in an online aversive learning system and in generating associated motivated behavior in humans.high-resolution fMRI | conditioned behavior | pallidum L earning which stimuli predict positive and negative outcomes, and thus should be approached or avoided, respectively, is central to an organism's ability to survive. Midbrain dopamine neurons respond to both unpredicted rewarding stimuli and to cues previously paired with rewards (1), consistent with behavioral approach toward those cues. As a counterpoint to these reward-related signals, neurons in the lateral habenula (LHb) of nonhuman primates respond to previously learned stimuli predicting the delivery of punishments and the omission of rewards, whereas they are inhibited by stimuli that signal upcoming rewards (2). These studies in nonhuman primates have concentrated on well-learned stimuli, and so have forsaken the opportunity to study the details of dynamic adaptation in the habenula. However, in many real-world scenarios, organisms learn about the motivational value of novel cues in their environment gradually, one exposure at a time, which raises the question as to whether the habenula plays a role in encoding the dynamically changing motivational value of cues that predict negative events.Dynamic learning from aversive events permits the rapid experience-dependent updating of behavior, for example, the automatic suppression of approach, which is a characteristic of aversive conditioning (3). The LHb receives inputs from the globus pallidus (4), and its excitation inhibits midbrain dopamine neurons via the rostromedial tegmental nucleus (2). This position as a hub between corticolimbic networks and midbrain monoaminergic nuclei provides a means through which positively or negatively valenced stimuli can modulate motor output, leading to the hypothesis that the habenula plays a critical role in motivated behavior (5).Studies using temporally precise optogenetic stimulation of the LHb in rodents provide convincing evidence that the habenula drives behavioral suppression (6). This structure has been suggested as a novel target for de...