For patients with certain diseases, blood stem cell transplants can be life-saving. But a transplant is likely to be successful only if the immune systems of the donor and recipient are a close genetic match. Human immune systems are extremely diverse and a patient's chances of finding a matching donor would be small without the ability to search a very large number of possible donors. For this reason, most advanced nations maintain large registries of potential donors who have offered to donate stem cells if they happen to be the best available match for a patient needing a transplant. In recent years, transplantation of stem cells retrieved from umbilical cord blood of newborn infants has emerged as an alternative technology. In addition to maintaining adult donor registries, many nations now collect and store inventories of cord blood to be made available when and if a matching patient appears. Institutions have arisen that allow transfers of adult donations and of cord blood units across international boundaries. Since no one knows in advance whether they will need a transplant in the future, and since larger registries and banks increase the probability of a match and hence of survival, the registries and cord blood banks present an interesting example of a pure public good with benefits that are fairly evenly dispersed. This paper explores the gains in survival probability that arise from increased registry sizes and uses ``value of statistical life'' methods to estimate expected benefits and compare them to estimated costs of additions to the adult registry and to the cord blood banks. Our analysis takes account of international flows of blood stem cell materials and compares outcomes when nations act separately in their own self interest with those in which they act cooperatively. Our results suggest that for the world as a whole and for large countries, the sum of marginal benefits of an increase in either the adult registry or the cord blood registry exceed marginal costs. However, marginal benefit-cost ratios for the adult registry are greater than those for the cord blood banks, which suggests that to the extent that these two sources of life saving compete for public funds it may be preferable to prioritize expansion of the adult registry over cord blood banks.