This article presents a methodology for identifying critical links in global resource supply chains by tracking resources from their extraction in one region of the world economy through their embodiment in intermediate products in the same and other regions to eventual embodiment in final goods. We build on previous work that applied an absorbing Markov chain (AMC) to results obtained using an input-output (IO) model of a single region to define a resource-specific network within that economy. In the absence of model calculations, the AMC can also be applied to standard IO data for a past year. This article first generalizes the analytic framework from a single region to the important case of the global resource-specific network. This network typically includes cycling of embodied resources between sectors not only within each economy, but also among regions, as subsequent rounds of intermediate products are traded. Next, we refine that analysis to exhibit a crucial subnetwork, the resource end-use network, which only tracks the portion of the resource that ends up embodied in a specific final product in a given region. Finally, we develop techniques to distinguish key branches of these networks and provide detailed insights about the structure of global resource dependence. A numerical example is applied to results of scenario analysis using an IO model of the world economy. Two alternative scenarios are compared. In each scenario, embodied resources are carried over specific branches of a global network in three regions using three resources to produce four goods.