We analyze how children’s disability affects intra-household investment decisions. By means of a general preference model, we show that variation in family size and health conditions can be used to infer whether parents are averse to inequality in the distribution of quality among their children or if, instead, they care more about efficiency. In particular, we exploit the fact that parents of only children cannot possibly exhibit inequality aversion. We apply our identification strategy to Mexican cross-sectional data and find evidence that parents are inequality averse. Specifically, our results show that inequality aversion induces an average increase of 0.7-0.8 years of schooling for disabled individuals when non-disabled siblings are present. We also show that the effect differs by child’s gender with large and significant effect for males and not statistically significant effect for females. While parental inequality aversion does not close the schooling gap between disabled and non-disabled males, its estimated effect is economically relevant, as it represents about 13-15 percent of the disability gap in education, which amounts to 5.3 years of schooling in Mexico.