Islands host exceptionally high levels of endemism compared to mainland regions and are subject to disproportionately high rates of extinction and imperilment. Therefore, the protection and preservation of taxonomic units that are endemic to islands is a key component in mitigating the loss of global biodiversity. However, determining what is “endemic” on islands can be challenging. Conservation units are commonly delineated based on genetic divergence at neutral loci (e.g., genetic differentiation at microsatellite loci or reciprocal monophyly based on mitochondrial genes). Island populations of nonvolant species are expected to meet this criterion, regardless of adaptive differences, due to geographic isolation, founder effects, and small effective population sizes. We therefore argue that the delineation and management of island endemic populations should not be based on neutral genetic divergence and reciprocal monophyly alone. Instead, we recommend identifying island populations that have genetically based adaptations to their unique environments. A comprehensive framework specifically designed to delineate evolutionarily significant units (ESUs) on islands should be based on metrics of both neutral and adaptive genetic divergence. The California Channel Islands host several taxa considered to be endemic, and we highlight 2 case studies to illustrate how this framework can be applied. This approach can be applied broadly to continental islands and island archipelagos, enabling conservation practitioners to use an objective framework to prioritize units of biological diversity for management.