Over the last decade, African courts have played an important role in the conductof free and fair elections. Both in Malawi and Kenya, the Supreme Courts nullified the presidential election of the incumbent. These rulings challenge the conventional wisdom that courts in hybrid regimes always support power holders despite blatant evidence of electoral fraud. Furthermore, these sparks of judicial resistance deserve our scholarly attention as they have the power to start the democratization process. Hence, this dissertation seeks to understand the conditions under which courts side with the opposition. To do so, I develop a novel theory of judicial resistance. I argue that institutional reforms incentivize courts to resist the incumbent’s pressure and to rule in favor of the opposition during electoral disputes under two conditions. First, institutional reforms must reinforce de jure independence by enacting effective legal and constitutional mechanisms that prevent the executive branch from undermining the separation of power. Second, institutional reforms must mobilize judicial support networks by granting them the tools to engage in strategic and repeated litigation. I test my argument by leveraging original qualitative and quantitative data from Kenya and Senegal, which are key cases because they allow me to challenge the assumption that courts’ behavior is shaped by levels of democracy. This dissertation makes important contributions. First, by disaggregating judicial institutions and identifying the different pathways that are conducive to judicial resistance or judicial manipulations, I show that courts in hybrid settings are not inherently weak and can become sites of contestation during the electoral period. Second, I show how elections can facilitate – or undermine – democratization by shedding light on the understudied role that courts play in the process. Finally, I make an empirical contribution by creating the first dataset on electoral disputes in Africa.