Plant transformation and regeneration systems have become indispensable parts of gene discovery and functional characterization over the last two decades. Adoption of transformation methods in studies of plant adaptation to natural environments has been slow. This is a result of poor genomic knowledge and inefficient transformation systems for species dominating terrestrial ecosystems, and logistical difficulties in conducting field tests of genetically engineered organisms. In trees, where long generation cycles, high background polymorphism, large sizes and outcrossing systems of mating make production of near-isogenic lines and large experiments difficult, transformation is an attractive alternative for establishing direct linkages between genes and adaptively significant phenotypes. Here, we outline the capabilities, challenges, and prospects for transformation to become a significant tool for studying the ecophysiological adaptation of trees to the environment. Focusing on poplars (genus Populus) as model system, we describe how transformation-based approaches can provide insights into the genes that control adaptive traits. The availability of the poplar genome sequence, along with its large expressed sequences tag (EST) databanks, facile transformation and rapid growth, enable reverse genetic approaches to be used to test virtually any hypothesis of gene function.