Although reproductive health advocates consider family planning programs the intervention of choice to reduce fertility, there remains a great deal of skepticism among economists as to their effectiveness, despite little rigorous evidence to support either position. This study explores the effects of family planning in Ethiopia using a novel set of instruments to control for potential non-random program placement. The instruments are based on ordinal rankings of area characteristics, motivated by competition between areas for resources. Access to family planning is found to reduce completed fertility by more than one child among women without education. No effect is found among women with some formal schooling, suggesting that family planning and formal education act as substitutes, at least in this low-income, low-growth setting. This provides support to the notion that increasing access to family planning can provide an important, complementary entry point to kick-start the process of fertility reduction.