My dissertation examines the geography of state building in 20th century Africa and assesses the effects of local state capacity on development and conflict. I argue that rulers invest in building their states to increase the revenue they extract from society. Trading taxes for public services, rulers' profits are constrained by their bargaining power and transaction costs. They tilt both in their favor by investing in direct or indirect rule over internal competitors and by building transport infrastructure that connects them to their people. To maximize returns, rulers target their investments to local socio-geographic environments. The result is spatial variation in state capacity, which affects local development and conflict. \ \ The first empirical part of the dissertation analyzes African states' investments into local state capacity. To analyze where colonizers increased their bargaining power, I present new data on local direct and indirect rule in French and British colonies. These data show that British governments crushed precolonial polities half as often as their French counterparts, who commanded more administrative resources and mirrored direct governance in metropolitan France. The data also show that the British ruled indirectly over precolonial states, but directly over societies without centralized institutions to co-opt. I then examine where rulers reduced their transaction costs by investing in transport infrastructure. To this end, I gather road network data from British colonial reports (1900--1950) and use machine learning to digitize road maps of post-colonial Africa (1966--2016). Highlighting the extractive roots of local state reach, I find that resource-rich areas have featured more roads and better connections to ports and capitals throughout the century. \ \ The second empirical part examines the effects of indirect rule and local state reach on development and conflict. As indirect colonial rule increased the bargaining power of local populations, indirectly ruled citizens received more public services in return for producing resources taxed by their governments, in particular cash crops. Analyses that exploit variation in indirect rule across neighboring ethnic groups and within groups split by French-British borders bolster this result. Citizens also profit from expanding trade with the government that comes with lower transaction costs. Assessing the impact of local changes in travel times towards administrative capitals, I find generally positive effects on local education, infant mortality, and nightlight emissions. Lastly, transport infrastructure allows governments to keep the peace but can also facilitate rebel mobilization. Drawing on credibly exogenous variation in simulated road networks, I find that roads that connect ethnic groups with national capitals curb conflicts, while roads that connect groups internally increase the risk of violence. \ \ My dissertation adds to the literature by analyzing the extractive roots of local state capacity in Africa, highlighting the interaction between colonial and precolonial institutions, and focusing on the trade between states and their citizens to explain local development and conflict. New data on local state capacity in 20th century Africa enable these contributions.