Diverse methods are available for evaluating benefits and costs of rural research projects. They have been developed in response to a felt need for information in a highly uncertain environment. These formal evaluation schemes are compared in an attempt to show whether any offer useful guidelines for rational allocation of research funds. The outcome is a series of conjectures on the level of effort to devote to research evaluation, and hence on the techniques which should be used. In most circumstances encountered in Australian rural research the optimal type of evaluation would be relatively unsophisticated.