This thesis presents an original theory of authoritarian party and legislative institutions in Africa, tracing their trajectories from an early period of regime consolidation through subsequent episodes of socio-economic and institutional change. Contra a dominant rational choice literature on authoritarian institutions, as well as an Africanist analysis of "neo-patrimonialism", I demonstrate how institutional variation reflects differences in the underlying distribution of power across African regimes. I argue, first, that variation in legislative strength and assertiveness is a function of the institutional strength and cohesion of ruling parties. The institutional make-up of these parties varies, in turn, depending on the early strategies of "politicized accumulation" and patronage distribution deployed by authoritarian leaders, as well as subsequent patterns of economic change. The legislature remains more marginal and subservient where authoritarian leaders work to centralise wealth accumulation, control patronage distribution and build up party institutions to channel and constrain elite contestation. By contrast, parliaments assert themselves where more diffuse patterns of accumulation fuel patron-client factionalism, undermine party cohesion and turn the legislature into an arena for intra-elite bargaining. Beyond analysing when and how parliaments strengthen, I also reassess the significance of a more assertive legislature, particularly its implications for distributive outcomes. My explanation of institutional variation yields fresh insights regarding whose interests a stronger legislature is likely to represent. Once we appreciate the role of elite contestation in driving legislative activity, it follows that parliamentary interventions tend to serve elite interests, reinforcing an existing wealth inequality. To demonstrate this argument, I use a combination of within- and cross-case analysis, drawing primarily on the Tanzanian and Ugandan cases with further reference to Kenya and Rwanda. I adopt a process tracing methodology to assess the validity of my causal argument and, for evidence, rely primarily on qualitative data drawn from elite interviews, archival work and observation of relevant party and legislative meetings.