When is it acceptable to buy a vote? Or rig an election? What do voters demand of their candidates, and what counts as good leadership? Through what process is legitimate electoral activity defined and contested? Drawing on three years of research on Kenya, Ghana, and Uganda, this paper investigates the moral economy of elections in Africa. In adopting this focus, we are explicitly seeking to go beyond analysis that depicts African elections as being immoral or amoral. Against such approaches, we demonstrate that one cannot fully understand electoral processes without first making sense of moral debates about what does and does not constitute acceptable electoral behaviour in different contexts. In using the term moral economy to describe these debates, we are not trying to evoke a singular, widespread set of behavioural rules that mobilise popular involvement in elections; nor are we suggesting that this is an inherently conservative force. Rather, we are suggesting that elections – by their nature – provide a powerful focus for speech and behaviour that makes claims to virtue, and /or demands virtue of others. The nature of virtue is contested, and people may be pulled in different ways by competing moral registers. The moral economy is manifest in the contingent outcomes of these debates, in apparent contradictions of behaviour, in the inability of civic education campaigns to eradicate gift-giving and in the willingness of politicians and their campaigners to flout the very laws to which they appeal for justice. Moral economy provides a way to think about these contradictions, and about the ways in which power is both reproduced and contested. It is the consequent messiness, and contradictions, which make African elections so vivid – and so resistant to interventions which champion a singular model of electoral virtue.