Since the 1970's, the worldwide capacity of genebanks for ex situ conservation of crop genetic resources has increased greatly. This has increased the accessibility of landraces and wild and weedy relatives to crop breeders; in situ conservation, though essential, is not an efficient means of furnishing genebanking services. But utilization of genebank resources has not kept pace. The set of popular cultivars in major crops is typically rather small, and their ancestry encompasses only a small fraction of the genetic diversity currently available in other cultivars. Discussions of farmers' rights that focus on compensation for current incorporation of farmers' varieties in new cultivars have diverted attention from the question of why so little of the newly accessible genetic diversity is currently being utilized by public and private breeders. To optimize the future provision of genebank services, research is needed on the costs of genebanks, the market for their services, the use of genetic resources by breeders, and the implications of recognition of farmers' rights, evolving intellectual property rights, continued funding problems and developments in biotechnology.