Why do people cooperate? A central explanation involves reciprocity: if we interact repeatedly, then I have an incentive to cooperate with you so that you will cooperate with me in the future. The same logic applies to punishment: if you are selfish to me today, I have an incentive to punish you so that you will not do so again in the future. But why, then, do people cooperate and punish even in one-shot anonymous interactions, where no such future benefits exist? As a theoretical framework for answering this question, I have proposed the “Social Heuristics Hypothesis,” which posits that we internalize typically successful behaviors as intuitive heuristics for social interactions. Deliberation can then shift our behavior toward what is most advantageous is the specific situation we face at any given time. Because most of our important interactions (e.g. those with our co-workers, friends, and family) are long-term rather than anonymous and one-shot, I argue that we intuitively apply a ‘repeated game’ heuristic: our intuitions support strategies which are payoff-maximizing in repeated interactions. In this talk, I will present this theoretical framework, and then validate its predictions with a series of economic game experiments that use a dual process framework to investigate the cognitive basis of cooperation and punishment. I will show that in one-shot cooperation games with strangers, cooperation is typically intuitive, except for people whose daily life interactions partners are non-cooperative (such that cooperation would not be advantageous even in a repeated game) or who extensive prior experience with economic game experiments (and thus have learned to distrust their intuitions). Similarly, paying to punish someone that has wronged you in a one-shot anonymous interaction is favored by intuition. Thus our intuitions implement strategies that are successful in the repeated interactions of daily life.