Do variations in land ownership affect people's democratic participation? Quantitative, cross-country research on this topic suffers from the non-comparability of regulatory systems and cultures, and the use of crude indicators to identify participation. This study attempts to overcome these methodological problems, by employing indicators of procedural and substantive participation in a structured, diachronic comparison of qualitative data from five sites in China-an authoritarian state, which, however, requires residents of urban communities and villages to participate in 'self-government'. It examines whether and why changing land from collective ownership to state ownership, and residents' compensated acquisition of cash and secure, fungible assets, strengthens or weakens participation in self-government. In the research sites, collective land ownership is found to stimulate participation in self-government. Transformation of the land to state ownership and people's acquisition of private property weakens participation. The robust results of the study support the direction of a causal argument that collective land ownership is conducive to democratic participation. These findings imply that scholars and policymakers should consider the potentially adverse political consequences of changing land ownership. A further implication is that, absent substantial political reform, an urbanized China might be less rather than more democratic at the community level.