This memorandum is concerned with the allocation of resources to biomedical research. In particular, it asks the question: How much should the country spend on such research? A simple mathematical model of individuals' preferences is proposed that gives reasonable predictions to qualitative questions posed. In lieu of obtaining sample data about consumer preferences, the model is used to determine what biomedical research is worth to consumers. The answer is that the probability of achieving future advances appears to be sufficiently low to warrant the conjecture that the nation is already spending more for biomedical research than it is worth to consumers. This revision expands the original by explicitly treating uncertainty rather than assuming it away. In addition, the first version assumed, incorrectly, that the important parameter a could not be greater than one. This revision corrects that assumption. Certain other less important errors also have been corrected.