Because parental care is expected to depend on the fitness returns generated by each unit of investment, it should be sensitive to both offspring condition and parental ability to invest. The Trivers-Willard Hypothesis (TWH) predicts that parents who are in good condition will bias investment towards sons, while parents who are in poor condition will bias investment towards daughters because high-quality sons are expected to out-reproduce high quality daughters, while low-quality daughters are expected to out-reproduce low quality sons. We report results from an online experiment testing the Trivers-Willard effect by measuring implicit and explicit psychological preferences and behaviorally implied preferences for sons or daughters both as a function of their social and economic status and in the aftermath of a priming task designed to make participants feel wealthy or poor. We find only limited support for predictions derived from the TWH and instead find that women have strong preferences for girls and men have preferences for boys.