The importance of the agricultural sector in the economic development process is well known. Improvements in agricultural productivity are often found to spill into other areas of a developing economy, potentially improving the standards of living of urban and rural workers alike. Given the importance of this sector, accurate measures of total factor productivity (TFP) across countries can be helpful in identifying conditions, institutions or policies that promote agricultural development. In this article, we estimate TFP growth in agriculture for a panel of 39 sub-Saharan African countries from 1961 to 2007. We also develop a set of development outcome measures theoretically consistent with strong agricultural performance to serve as external validation of our results. We find that three estimation methods (stochastic frontier, generalised maximum entropy, and Bayesian efficiency) generate relative rankings that are consistent with the development outcome measures, providing external validation of the methods. However, the data envelopment analysis approach performs poorly in this regard.