Despite growing urgency to stem biodiversity loss and support human health and wellbeing, the resources available for conservation remain limited. Resource constraints necessarily demand that decision-makers and practitioners be strategic to ensure strong returns on investments. Strategic action is most likely to occur when organizations use scientific information to set priorities and rigorously evaluate project impacts. This dissertation examines priority-setting and evaluation within the context of three conservation challenges – global prioritization of locations to protect biodiversity and ecosystem services, community and state-led efforts to conserve forests in Madagascar, and spatial planning for co-benefits for ecosystem services and biodiversity in the U.S. I also reviewed the literature and developed recommendations intended to build capacity for conducting conservation impact evaluation.In the global study, my collaborators and I identified global spatial priorities for biodiversity and ecosystem services (termed ‘nature’s contributions to people’, NCP) and evaluated where conservation and development pressures might conflict. We conducted a global-scale spatial optimization using data on 10 NCP and 26,709 species of birds, mammals, reptiles, and amphibians. To identify areas of potential conflict with development, we overlaid resulting conservation priorities with data on areas of high suitability for agricultural expansion, renewable energy development, oil and gas production, and urban expansion. We found that approximately half (44-49%) of global land area, excluding Antarctica, should be conserved through protection or sustainable management to provide 90% of current levels of NCP while also meeting species representation targets. Notably, the land area required to achieve these benefits is consistent with commitments made by governments under the Kunming-Montreal Global Biodiversity Framework to conserve 30% of global lands and waters by 2030, as well as proposals by scientists to conserve “half Earth” for biodiversity and climate goals. However, our findings also point to potential conflict, as 37% of priority areas overlapped with areas of high development potential, especially with renewable energy (solar, in particular) and commercial agriculture. The potential for conflict is amplified by the fact that only 18% of priority areas are currently protected. Thus, the conservation community will be challenged to identify a portfolio of interventions and management approaches that can accommodate some level of human activity while also safeguarding biodiversity and ecosystem services. Our work illustrates how the complexity of socioecological systems and their contributions to humanity should be explicitly considered when setting priorities. In our U.S.-focused project, we investigated how protecting ecosystem services may accrue co-benefits for bird conservation. We quantified co-benefits by examining the spatial convergence between national-scale spatial priorities for eleven ecosystem services (ES) and populations of 479 species of birds, using data from the participatory science project, eBird. For each species, we calculated the percentage of its U.S. population that occurred within ES priority areas and summarized the information by biome communities (e.g., forest, grassland, aridland). Given high interest in carbon for mitigating global climate change, we separately examined spatial overlaps with priority areas for ecosystem carbon storage. ES priority areas contributed meaningfully – and more than expected by chance – to the populations of over half of U.S. bird species, including so-called “Tipping Point” species that urgently require conservation action. Co-benefits were greatest for forest birds, such as Cerulean Warblers (Setophaga cerulea) and Eastern Whip-poor-wills (Antrostomus vociferus), but also were evident for certain grassland, wetland, and aridland species. Other species were poorly represented within ES priority areas, and this included some declining species, such as Snowy Owls (Bubo scandiacus) and Pinyon Jays (Gymnorhinus cyanocephalus). Our findings speak to the need for caution when generalizing or assuming about the degree of co-benefits that can be realized in specific projects. Our Madagascar case study in forest conservation focused more squarely on impact evaluation, which seeks to measure the effectiveness of conservation interventions. Much of the world’s biodiversity is concentrated in tropical nations, many of which struggle with recurring political instability. We sought to understand which interventions are most effective during a crisis, with a focus on Madagascar, a global biodiversity hotspot. We investigated the effects of a particularly severe and prolonged political crisis (2009-2014) on the relative performance of two types of conservation interventions: community forest management (CFM) and protected areas administered by Madagascar National Parks (MNP) in terms of their ability to reduce deforestation. Using remotely sensed data on forest cover and a range of spatial and temporal covariates, we conducted statistical matching and an event study design. Used in combination, these two methods seek to control for time-invariant and time-variant confounding factors which can influence deforestation performance, such as systematic differences in location of CFM and MNP forests, and temporal events such as cyclones and changes in commodity prices. Results showed that, although both CFM and MNP lost forest cover before, during, and after the crisis, annual rates of deforestation were higher in CFM than in MNP across all years in our study period (2005-2020) and accelerated at the end of the crisis. After controlling for confounding factors, however, we found no difference in performance during the crisis years (2010-2013). This changed during the post-crisis period (2014-2017) when CFM experienced annual deforestation rates that were 1.7-2.4 times higher than equivalent areas within MNP. In other words, both CFM and MNP performed poorly during and after the crisis, but CFM performance deteriorated significantly in the post-crisis period. Results suggest that when the crisis subsided and deforestation pressures intensified, community-based conservation was less effective than state protection. Our work underscores the importance of improving the resilience of conservation efforts to political and economic shocks. Lastly, we explore lessons from the literature on conservation impact evaluation and offer recommendations for bridging the gap between impact evaluation research and practice. Globally, billions of dollars are invested in nature conservation every year, but the impact of those funds on biodiversity or people are still poorly understood. A growing number of conservation impact evaluation studies, including our analysis in Madagascar, seek to tease apart the causal effect of conservation from the complex web of social, economic, and biophysical factors that influence outcomes. Such research can help governments, businesses, philanthropic institutions, and conservation organizations measure the effects of conservation investments. Impact evaluations can be expensive and time-consuming, but without them one risks allocating scarce resources on ineffective strategies. Given these tradeoffs, we argue that that impact evaluations are most valuable for projects that (1) present a high risk of failure or unintended outcomes, (2) require large budgets, (3) occur across large scales, (4) rely on new or unproven approaches or (5) assume additionality. We offer practical recommendations for projects that decide to undertake impact evaluation, including building evaluation into project design from the start and jointly fundraising with researchers to avoid diverting funds from implementation. We also suggest ways funders can contribute, including requiring credible evidence for causal claims, setting realistic expectations, and rewarding learning rather than results. In summary, my dissertation examines how conservation practice can be improved through forward-looking planning as well as retrospective evaluation, and at scales ranging from global to national. While each chapter provides unique insights, a unifying theme is improving the practice of conservation for the benefit of all species on the planet, including our own.