This paper examines the effect of birth order on educational attainment in Lesotho. Using family fixed effects models, I find robust negative birth order effects on educational attainment. These results are in sharp contrast with the evidence from many developing countries, but are consistent with that from developed countries. Further, these birth order effects are pronounced in large families, and families with first-born girls, which suggests presence of girls' education bias. Turning to potential pathways of these effects, I find that they are not propagated through family wealth, but mainly through birth-spacing. These results are robust to different sample restrictions.