There have been several important cases of movement-based parties that rose rapidly in popularity and were able to attain national power in new democracies. Existing theories predict these parties will become increasingly top-down organizations designed to preserve and enhance the power of party elites, a trend that is usually aggravated when parties govern nationally. The Bolivian MAS deviates from this conventional wisdom, as it has followed a remarkably different organizational trajectory that has facilitated grassroots impact and constrained elite control. Through a within-case comparative examination of the MAS, this study identifies necessary conditions and explains mechanisms facilitating this outcome in the crucial areas of candidate selection and national policy-making. The study finds that a set of historical legacies traceable to a party's origins and structural elements associated with the density of civil society heavily affect power distributions within governing movement-based parties. Both elements can facilitate the emergence of opposition among allied groups that can check power from within and keep open channels for agenda setting from below. The realization of this potential, as this study argues, depends heavily on the organizational strength, unity, and mobilization capacity of allied groups in civil society. The analysis reveals that movement-based parties are remarkably flexible organizations whose boundaries with allied groups in civil society tend to be fluid and empirically blurred. The empirical basis for this argument derives, first, from a wealth of qualitative data collected in Bolivia, where I conducted twelve months of fieldwork in different regions of the country. During that time, I conducted over 170 in-depth interviews with party elites at the national, state, and municipal levels, as well as with a wide variety of civil society actors, including union leaders, activists, opposition politicians, and others. Second, cross-national comparisons with the experiences of the Brazilian Workers' Party (PT) and the Uruguayan Broad Front (FA) improve the overall evidentiary base of this study. They also help to further support the theoretical claims about the importance of historical and structural factors for shaping the degree of power concentration in movement-based parties.