Large-scale anthropogenic disturbances by fire, extractive industries or other uses of wildlands is driving the effort to restore natural communities quickly before excessive degradation by secondary factors such as erosion or invasion by exotic species makes recovery virtually impossible. Until recently, ecological restoration has aimed at replacing damaged vegetation with locally collected plant materials so that suitability to the local environment can be assured. In practice, local collections may be unavailable for large scale restoration efforts and materials that may be suitable today may not remain suitable over the life span of the species due to rapid climate change. A concerted scientific approach is therefore needed to match genes and traits with environmental conditions to establish populations that are viable in the long run. Instrumental in this research agenda has been the implementation of multiple common garden experiments, in which ecotypes collected from the full range of a given species are each grown under a range of environmental conditions. However, this ostensibly straight-forward approach for selecting traits or genotypes on the basis of growth and survivorship in a given environment also has potential pitfalls. We asked our speakers to focus their remarks on three basic questions: • Which relevant adaptations do we overlook by working with greenhouse-raised transplants and what are the implications for restoration? • How do we interpret short-term responses to environmental fluctuation in terms of the climate adaptation of populations? • What is the best-practice protocol for interpreting the results of multiple common garden experiments for restoration?