Scholarship on contemporary revolutions suggests that a revolutionary coalition of socially, ideologically, and ethnically diverse citizens is necessary for toppling durable autocracies. Yet what are the ramifications of such "negative" revolutionary coalitions for politics once the revolution ends, and the transition to democracy begins? In Tunisia, the principal contestation of the “transitional” period has occurred not between revolutionaries and old regime supporters, but between dueling factions of the revolutionary coalition that unseated Ben Ali in 2011. Emergent theoretical literature on the “urban civic revolution” suggests that rapidly assembled, diverse revolutionary coalitions are prone to post-revolutionary fractionalization – and that this fractionalization may in turn drive patterns of political and social contestation during the onset years of a new political order. During the Tunisian political crisis in August 2013, co-participants in the 2011 revolution found themselves in opposing protest groups, each side claiming to defend and uphold the revolution. What can the demographic, behavioral, and attitudinal characteristics of these groups tell us about how – or more specifically, along which political lines – the revolutionary coalition has broken down? Using an original dataset of pro- and anti-government protest surveys collected from Tunis in August 2013, this paper tests a series of empirical hypotheses regarding the dynamics of revolutionary coalition fragmentation in Tunisia. I argue that a research design comprising protest surveys of competing demonstrations lends empirical insight into the nature of emergent societal coalitions in transitional states. I provide the first evidence that both of these competing coalitions are indeed characterized by strong participation in the 2011 revolution, and show that fragmentation of this erstwhile revolutionary coalition has occurred along ideological lines concerning the role of Islam in governance, rather along the lines of social class, as predominant theories of revolution and democratization would suggest. I furthermore show that personalized evaluations of post-revolutionary progress, particularly in the economic arena, perform an important sorting function between pro- and anti-government groups. Finally, I develop and test a mechanism of protest mobilization specific to post-revolutionary contestation, namely the articulation of divergent revolutionary narratives corresponding with the predominant characteristics of each group.