Lab-in-the-field experiments occupy a middle ground between laboratory studies and full-scale field experiments. This talk explains the rationale for using such designs to study theoretically relevant populations, focusing on experiments conducted with CEOs in China. The results show systematic deviations from standard economic predictions, including higher levels of trust and cooperation, that are nonetheless patterned and meaningful. The findings underscore the importance of subject selection for theory testing and demonstrate how lab-in-the-field experiments can improve both explanatory relevance and empirical validity.