We argue that classic economic games and their more recent extensions should continue to play a role in fieldworkers’ methodological toolkits. Economic games are not replacements for observational and self-report studies of behavior, but rather complements to them: While observational and self-report data measure individuals’ behavior subject to the constraints of cultural institutions, competing demands on their resources, and even self-presentation bias, economic games can be designed to measure comparatively unconstrained individual preferences, or to selectively introduce constraints, providing insight into how individuals would behave under certain conditions if they had the opportunity. By using a combination of experiments, observation, and self-report, anthropologists, economists, and psychologists can continue to improve their understanding of how preferences translate into “real world” behavior, and how “real world” constraints influence preferences, across diverse human societies.