AbstractA practical means of long term genetic conservation in forest trees is to establish natural (in situ) populations (i.e. gene resource management units, GRMUs) or ex situ plantings as gene conservation reserves. Results from pollen contamination studies in conifer seed orchards, however, indicate that gene flow in such reserves could be extensive. Although gene flow can be beneficial in terms of introducing new genetic variants, immigration of genes from domesticated populations is likely to reduce total genetic diversity within reserves and potentially lower their fitness. The prudent position on gene flow in reserves, therefore, is to limit it as much as possible. Pollen gene flow can be eliminated in ex situ plantings by controlled mating. If controlled mating is not feasible, applying pollen management techniques to increase pollen production within the plantings relative to external sources can minimize gene flow. Strategies for reducing gene flow in GRMUs are to make reserves as large as possible, include buffer zones around reserves (perhaps planted with an exotic species), and to assure that natural regeneration, or seed collection for artificial regeneration, occurs in heavy seed-crop years. Gene flow in forest trees is still poorly understood and the effectiveness of various approaches to limit gene flow in reserves, especially in GRMUs, cannot be quantified with any accuracy. Thus, research on gene flow in forest trees should receive high priority.