The importance of connectivity for species conservation has resulted in myriad attempts to identify corridors linking habitat patches and conservation areas. However, making smart decisions for restoring connectivity requires information beyond simple maps of corridors. Here, we combine land-parcel cost estimates with a new analytical approach that pinpoints where barrier removal can best improve connectivity to develop a return-on-investment framework for connectivity restoration. An iterative series of barrier analyses followed by simulated restorations allowed us to incorporate cumulative effects of previous restoration decisions, which can alter the potential value of future restorations. To demonstrate our approach, we identify specific sites that, if restored, would most increase habitat connectivity for the Washington ground squirrel (Urocitellus washingtoni). The analysis was performed in two ways: first without consideration of economic costs and then again, explicitly incorporating the costs of land purchase and restoration. We found that accounting for land costs could reduce overall restoration costs by 55% and increase the area of land restored by 30%. These methods can be tailored to meet different connectivity objectives and can complement reserve design selection methods that prioritize conservation of contiguous natural areas.