This paper uses a cross-sectional approach to analyze the impacts of climate change on animal husbandry and the way farmers adapt. The study is based on surveys of almost 5000 livestock farmers across ten countries in Africa. A traditional Ricardian regression finds that the livestock net revenues of large farms in Africa are more sensitive to temperature than those of small farms. Cross-sectional analysis also reveals that large farms (but not small farms) have fewer animals per farm in warmer places. Farmers tend to select beef cattle and chickens in cool climates and goats and sheep in hot climates. Using the Ricardian results and examining climate scenarios for 2060 and beyond, the net revenues of small farms are predicted to increase as much as 120% (+USD6 billion) but those of large farms are predicted to fall by 20% (-USD12 billion). The impact estimates in any given period also depend on the rainfall predictions. The results suggest that large livestock farms in Africa are more sensitive to temperature than small ones, primarily because of their dependence on cattle.