Under poor circumstances, co-residence of a grandmother is generally considered to be beneficial for children. Empirical evidence does not unequivocally support this expectation and suggests that the grandmother’s importance depends on the family’s circumstances. We study the relationship between grandmother’s co-residence and children’s schooling in sub-Saharan Africa under a broad range of circumstances. Results make clear that the effect of a co-residing grandmother varies but is almost always positive. Grandmothers over age 60 are most effective in helping their children. They (...) are particularly important for girls, and when the mother is deceased or not living in the household. Grandmothers are less effective in situations with few opportunities, as in very poor regions or in communities with few schooling opportunities. Our findings indicate that providing support to grandmothers should not be overlooked when designing policies aimed at strengthening the position of women and children in the sub-Saharan African context. (shrink)