Bone marrow transplantation has become the treatment of choice for certain hematologic diseases. However, only 30-40% of patients who might benefit from this procedure have a suitable family donor. Consequently, many centers have begun to explore the use of unrelated volunteer donors. Initial results have demonstrated the feasibility of this approach. As a result, a national effort has begun to recruit HLA-typed volunteers in order to establish a registry of individuals who would be willing to serve as bone marrow donors. This manuscript explores the potential impact of establishing such a registry. We find that a registry of attainable size could more than double the number of marrow transplants now being performed. However, even with a registry of enormous size, it will still not be possible to identify an HLA-matched donor for some patients.