Framed field experiments (experimental games) are widely used to assess factors affecting cooperation in management of the commons. However, there is relatively little attention to how details of the games affect experimental results. This paper presents qualitative and quantitative results from a framed field experiment in which participants make decisions about extraction of a common-pool resource, a community forest. The experiment was conducted in 2017–2018 with 120 groups of resource users (split by gender) from 60 habitations in two Indian states, Andhra Pradesh and Rajasthan. We test whether within-subject treatments (non-communication, communication, and optional election of institutional arrangements (rules)), remuneration methods, and design of the game board affect harvest behavior and groups’ tendency to cooperate. We also examine how characteristics of the community and players affect players’ choices in the game, with special attention to gender differences. Results reveal participants harvested substantially less than the Nash prediction even in the absence of communication, with men extracting less than women in both states. For male groups in both states, both communication and optional rule election were associated with lower group harvest per round, as compared to the reference non-communication game. For female groups in both states communication itself did not significantly slow resource depletion; but introduction of optional rule election did reduce harvest amounts. For both men and women in Andhra Pradesh and men in Rajasthan, incentivized payments to individual participants significantly lowered group harvest, relative to community flat payment, suggesting such payments stimulated deliberation among game players. Findings have methodological and practical implications for designing behavioral intervention programs to improve common-pool resource governance.