Lower fertility can translate into a more male-biased sex ratio if son preference is persistent and technology for sex-selection is easily accessible. This paper investigates whether financial incentives can overcome this trade-off in the context of an Indian scheme, Devirupak, that seeks to decrease both fertility and the sex ratio at birth. First, I construct a model where the effects of incentives are determined by the strength of son preference, the cost of children, and the cost of sex-selection, relative to the size of incentives. Second, I create a woman-year panel dataset from retrospective birth histories and use variation in the composition of pre-existing children as well as the state and the year of program implementation to estimate its causal effect. Devirupak successfully lowers the number of children by 1 percent, but mainly through a 2 percent decrease in the number of daughters. Faced with a choice between a son and only daughters, couples choose a son despite lower monetary benefits, and thus the sex ratio at birth unintentionally increases. A subsidy worth 10 months of average household consumption expenditure is insufficient to induce parents to give up sons entirely. Instead, Devirupak increases the proportion of one-boy couples by 5 percent. The proportion of one-girl couples increases only among the lowest socioeconomic status groups.