Despite the achievements of smallholders in Asia during the green revolution, there is scepticism that Africa’s smallholders — who dominate the farm area in most countries — can imitate this model and deliver agricultural growth. This paper assesses whether such pessimism is justified. Given the high transactions costs of hiring labour of farms, diseconomies of scale can be expected when labour is relatively cheap and abundant compared to other factors of production: which may explain the survey evidence that small farms often produce more per hectare than larger farms. In conditions of low development with relatively cheap labour, small units may have advantages over larger ones. The empirical record of performance of small and large farms in Africa is uneven and incomplete. Given the dominance of small farms in agriculture in many African countries, national data may be indicative of small farm performance. The record since the 1960s shows variable performance in agricultural growth through time and space, with slow growth in the 1970s followed by acceleration from the early 1980s. Even more striking is the difference in the performance of Northern and Western Africa compared to that of other regions of the continent. But the differences are not just regional: there is great variation across countries. While many African countries have a disappointing record of growth, thirteen doubled or more their production in the twenty years from the early 1980s onwards. These include countries where the bulk of output comes from small farms — Burkina Faso, Ghana, Mali, Niger, etc. Countries that have, or had, notable large-farm sectors such as Namibia, South Africa and Zimbabwe are well down the growth ranking. This proves little about scale since other factors are so much more important for agricultural growth, but it does show that to have an agriculture dominated by small farms is no obstacle to growth, and quite rapid growth at that.