Restoration is a major approach to reversing biodiversity loss worldwide. There has been considerable research on restoration methodology, but the theoretical basis for this work has derived primarily from applied fields such as agronomy and civil engineering.  When ecological theory has been applied to restoration, it has tended to focus either on restoring biological components of communities while ignoring ecosystem functioning, or on restoring ecosystem functioning without regard to communities.  We argue that the BEF perspective provides a unique opportunity to integrate these two views and improve restoration success in a number of areas.  First, restoring biodiversity may increase ecosystem functioning through “classical” BEF mechanisms such as complementarity or selection effects.  Second, BEF theory suggests that biodiversity may significantly increase the temporal stability of ecosystem function, and therefore service provision, in restored systems.  Third, evidence suggests that biodiversity may be necessary for the provisioning of multiple services in restored ecosystems because different species possess traits that affect different ecosystem functions.  We see an opportunity for economic analysis comparing biodiversity's benefits of risk management and provisioning multiple services to the costs of creating and maintaining diversity in restorations.  For BEF science to be applicable to restoration practices, more studies must take place in the restoration context (using realistic mixes of species, measuring functions related to ecosystem services), and better information on ecologically important traits is required.  Finally, there is a real need for greater effort on the part of scientists to ensure that results are conveyed to restoration practitioners.