Dams, road crossings, and water withdrawals extensively fragment rivers, and watersheds often contain hundreds or thousands of barriers, some of which no longer meet societal purposes. Accordingly, both conservationists and infrastructure managers are faced with the challenge of prioritizing barriers for repair, replacement, or removal. Candidate projects have been prioritized with dozens of methods, which span a wide range of spatial scales, data and analytical requirements, mathematical complexity, and capacity to reconcile multiple perspectives and objectives. We briefly review barrier prioritization methods from the perspective of a policy maker or manager who must balance realities of stochastic opportunities, conflicting priorities, and risk of infrastructure failure. After outlining common motivations for barrier prioritization, we present a menu of techniques ranging from large-scale, quantitative assessments to reactive, local response to failures. By clarifying the appropriate domain for each approach, this review informs the selection of prioritization methods for restoring riverine connectivity. K E Y W O R D S aquatic organism passage, dam removal, portfolio analysis, spatial prioritization, watershed planning 1 | INTRODUCTION Dams, road crossings, and other built infrastructure disrupt river and stream ecosystems yet serve a variety of important societal purposes including management of flood risk, provision of reliable water supplies, and transportation of people and goods. Alleviating ecological problems caused by infrastructure represents a critical conservation challenge. Specifically, these structures reduce aquatic connectivity, thereby limiting the ability of aquatic animals to reach important breeding, refuge, and feeding habitats; disrupting the transport of sediment; and altering river hydrodynamics (Pringle, 2001). The scale of infrastructure impacts on aquatic ecosystems is tremendous (e.g., Barbarossa et al., 2020; Vörösmarty et al., 2010). Although comprehensive geospatial databases do not exist, tens of thousands of large dams (Bellmore et al., 2016) and millions of road-stream crossings (Januchowski-Hartley et al., 2013