Ritual protocols aimed at rainmaking have been a recurrent sociocultural phenomenon across societies and throughout history. Given the fact that such protocols were likely entirely ineffective, why did they repeastedly emerge and persist, sometimes over millennia, even in populations with writing and recordkeeping? To address this puzzle, many scholars have argued that these protocols were not instrumental at all and that their practitioners were not really endeavoring to employ them to bring about rain. Here, taking advantage of the wealth of historical records available in China, we argue to the contrary: that rainmaking is best viewed as an instrumental, means-end activity and that people have always placed strong emphasis on the outcomes of such activities. To account for the persistence of rainmaking, we then present a set of cultural evolutionary explanations rooted in human psychology that can explain why people’s adaptive learning processes did not result in the elimination of ineffective rainmaking methods. We suggest that a commitment to a supernatural worldview provides theoretical support for the plausibility of various rainmaking methods and that people often overestimate the efficacy of rainmaking technologies because of statistical artifacts (some methods appear effective simply by chance) and underreporting of disconfirmatory evidence (failures of rain-making not reported or transmitted). The inclination to “do something” when a drought hits versus “do nothing” likely also plays a role and persists in the world today.