Maintaining and restoring connectivity are key adaptation strategies for biodiversity conservation under climate change. We present a novel combination of species distribution and connectivity modeling using current and future climate regimes to prioritize connections among populations of 26 rare species in New York's Hudson Valley. We modeled patches for each species for each time period and modeled potential connections among habitat patches by finding the least-cost path for every patch-to-patch connection. Finally, we aggregated these patches and paths to the tax parcel, commonly the primary unit of conservation action. Under future climate regimes, suitable habitat was predicted to contract or appear upslope and farther north. On average, predicted patches were nine times smaller and paths were twice as long under future climate. Parcels within the Hudson Highlands, Shawangunk Ridge, Catskill Mountains, and Harlem Valley had high species overlap, with areas upslope and northward increasing in importance over time. We envision that land managers and conservation planners can use these results to help prioritize parcel-level conservation and management and thus support biodiversity adaptation to climate change.