This dissertation examines factors affecting land policy performance in the Brazilian Amazon. It identifies the conditions and mechanisms influencing the success of weak bureaucracies in establishing property rights for rural smallholders and making land ‘legible’ (in public cadasters, survey maps, and land titles). I analyze how civil society and elites historically affected land legibility in Brazil and the contemporary functioning of Brazil’s national land bureaucracy at local levels.Social scientists posit that a strong state bureaucracy, a strong civil society interested in policy, elite consensus, or politically empowered peasants are necessary for land legibility and property rights formalization (Scott 1999:49; Emigh, Riley, and Ahmed 2019; Kingdon 2011; Murtazashvili and Murtazashvili 2021; Albertus 2021). Within weak bureaucracies, mission-driven bureaucrats would be necessary for organizational effectiveness (McDonnell 2020; Graizbord and De Souza Leão 2022), though the organizational mission is subject to partisan understanding and implementation (Merriman and Pacewicz 2022). I examine how civil society and elites influenced organizational effectiveness and land policy performance in surprising ways in two neighboring states—Pará and Amazonas—each more than three times the size of California and together containing most of the Amazon Forest. In Western Pará, the policy priority region, where interested elites and strong peasants pushed for land titling, the land bureaucracy was expected to produce high titling rates but, in fact, produced the lowest. In Amazonas, a non-priority state where elites were disinterested in rural land and peasants had little political power, the land bureaucracy unexpectedly titled most smallholders. Methodologically, my dissertation leverages subnational comparisons with process tracing, ethnography, and administrative data analysis. I conducted six months of fieldwork in land bureaucracies and rural communities. I completed ninety-seven semi-structured interviews with bureaucrats, civil society leaders, smallholders, and politicians. I collected administrative datasets with twenty-five right-to-information requests and archival data from land bureaucracies, civil society organizations, and local newspapers. The dissertation shows that in Western Pará, elites with different interests in land (politicians, agribusiness leaders, and public prosecutors) intervened and derailed the regional land bureau’s operation. While rural elites publicly supported private property rights, behind the scenes, they meddled with state legibility to advance illicit land uses. In response, environmental litigation unintendedly weakened bureaucracies further. External interventions imprinted partisan polarization onto bureaucrats and corroded organizational capacity. Few Pará smallholders received property titles. In Amazonas, a weak civil society and disinterested elites enabled the regional land bureau to function. Absent elite intervention, partisan bureaucrats increased organizational effectiveness by collaborating with smallholder civil society groups. Most Amazonas smallholders received property titles, especially in traditional communities. This dissertation contributes to ongoing development, state capacity, and climate governance debates by arguing that weak bureaucracies require shelter from (rather than support of) political elites to operate effectively. Bureaucrats’ polarization erodes organizational capacity, but moderate partisanship can foster collaboration with civil society to enhance organizational effectiveness. Furthermore, I argue that cooperative labor (rather than litigation and protests) enables local communities to achieve land legibility and environmental governance. My dissertation suggests that reorienting elite interests (or attention) away from local bureaucracies and facilitating cooperation with smallholders can implement state legibility to optimize land use and avoid climate catastrophe in Amazonia. Potential policy implications for sustainable development in environmentally strategic regions include international cooperation to strengthen local civil servants, incentives for non-predatory economic activities, and supporting environmental litigation strategies that target economic sectors rather than state institutions.