Adaptive control of cognitive behaviors in response to changes in uncertainty about future rewards is fundamental for survival. It is not surprising that uncertainty-related maladaptive behaviors, such as maladaptive risk seeking or avoidance, are observed in a wide range of psychiatric disorders. But, to date, the neural mechanisms of uncertainty-mediated risk seeking or risk avoidance are unclear. And, anatomically targeted treatments for many risk-related behavioral states have not been developed. The overarching goal of the current renewal is to uncover the neuronal and behavioral mechanisms of risky decision making under uncertainty. Aim 1 will uncover the mechanisms by which single neurons regulate risky decisions under uncertainty. Our previous work showed that the ventral pallidum (VP) transmits a risk signal before risky decisions. This raised crucial new questions. How does VP compute its risk signal? What is its function in decision making? We hypothesize that VP computes the subjective value of risk in order to govern risk- reward tradeoffs in decision making, and that decision making is controlled by a major recipient of VP projections - the lateral habenula (LhB). Aim 1 will (i) test among leading mechanistic accounts of risky behavior and assess how the subjective value of risk is governed by distinct forms of reward uncertainty and by reward timing, (ii) determine whether VP neurons' activity is necessary and sufficient to control decision making under uncertainty via these mechanisms, and (iii) determine whether and how VP risky decision related activity is reflected in the LHb. Preliminary data indicate that risky decisions arise due to a change in subjective value of risk mediated by uncertainty-sensitive neurons in the VP, and that representations of total subjective value ultimately guide choice through the LhB, downstream of VP. Aim 2 will shed light on the behavioral and neuronal mechanisms through which risk tolerance and value-based decision making are mediated by time. We found that monkeys' behavior closely resembled human decision making: monkeys are more willing to accept risks when they can resolve the resulting uncertainty early instead of having to live with the uncertainty for a prolonged time, and this risk tolerance was reflected in VP and LhB neurons. We hypothesize that timing information that is computationally necessary to control this risk tolerance is encoded in dorsal raphe nucleus (DRN) – a modulatory input to VP and LhB. Aim 2 will (i) characterize how neurons in DRN process the timing- and uncertainty-related variables that govern risk tolerance and (ii) transiently manipulate population activity in the DRN to assess the impact on decisions. Preliminary data show that DRN neurons encode information about time and uncertainty in a manner that is sufficient to control risk tolerance. The Aims offer an unprecedented opportunity to (i) understand the mechanisms of risky decision-making and (ii) study the n...