Information and communication technologies are blurring the boundaries between work and play. We present the first empirical investigation of gender gaps in virtual game economies. Analyzing big data sets from two major game economies, we find that player gender and character gender influence virtual wealth in different ways in different games. We conclude that this can be explained by different returns on female- and male-dominated play activities, that is, virtual pink- and blue-collar occupations. As the line between work and play increasingly blurs, researchers should track which occupations get to keep their conventional economic rewards, and which end up being remunerated in play money.