Botswana has figured widely as the exceptional African growth success story and has been frequently cited in scholarship that supports the view that African and other less developed economies are capable of rapid economic growth as long as the internal institutional framework and development policies are right. A shortcoming of the literature on African economic performance to date is that it has focused on the aggregate average growth rate, and has not taken the quality of the growth evidence into consideration. This article makes use of the official growth evidence taken from the published national accounts in Botswana to establish that there is reason to doubt the accuracy of the growth evidence on Botswana. It shows how the first decade of growth in particular is seriously biased upwards. After the official evidence and the national accounting methodologies have been analysed, several revisions of the African growth miracle become necessary. The ‘policy’ accounts have largely been informed by observing the recorded aggregate growth rate and have attributed the rapid growth to stylised facts about policies and institutions. A consideration of disaggregated growth rates allows a discussion of the causal coherence of the dominant explanations of rapid growth in Botswana, in particular, and in the divergent fortunes in the developing world, in general. This article argues that the growth miracle can only directly be attributed to economic policy if ‘good policy’ is defined as the absence of very bad policies.