Developed economies are largely based on impersonal exchange; trade does not require high levels of information about others' past behavior. Consequently, lack of trust, social distance, the possibility to break promises and other trade frictions all hinder economic activity. This project employs theoretical and experimental methodologies to study individual and group behavior in economies populated by individuals who do not know each other and may not trust each other. These strangers interact in pairs with randomly changing opponents whose identities and past behaviors are unknown. In each trade encounter, individuals can either cooperate or behave opportunistically. The sequence of encounters is indefinite, so many outcomes are possible. Theoretically, self-regarding individuals can overcome opportunistic temptations, and attain the best outcome, by adopting community-based punishment schemes. The research will pursue a systematic study of small and large laboratory economies based on these indefinitely repeated matching games. It will advance knowledge in the social and economic sciences by identifying behavioral elements and endogenously-arising institutions that are associated to the emergence, sustainability, and breakdown of cooperation among individuals who may neither know nor trust each other. New insights about cooperation will have applications to disciplines in which the study of social decision problems is an active area of research, such as political science, accounting, and sociology.