Gradual accumulation of evidence is thought to be a core component of many types of decisions, and neural correlates of it have been observed in multiple brain regions of primates. But whether any of these regions are necessary for the process, or what their specific roles might be, and how they work together to implement decision-making behavior, have never been elucidated. To facilitate addressing these issues, we developed a rat model of decisions guided by gradually accumulating evidence. Together with this rat behavior, we developed new computational methods to quantify different aspects of the decision-making process, developed a new method to estimate tuning curves for accumulation (firing rates as a function of the accumulating evidence), and to probe for necessity, performed the first pharmacological and optogenetic inactivations of four different brain regions thought to be involved in decision-making. I will describe how our results are beginning to give us a picture of the causal circuit logic of the system. In particular, a longstanding mystery has been the question of which brain region(s) are causally responsible for the accumulation process. I will describe how our most recent data begins to finally crack this problem, by indicating that the anterior dorsal striatum is the first known node of the neural circuit underlying accumulation.