Abstract. Numerous ranking formulae have been proposed in the literature for use in the evaluation and selection of R&D projects, but their rate of adoption has been relatively slow. This situation appears to be changing in the U.K. and a considerable amount of effort is being expended to improve the methods of obtaining estimates for the variables which appear in the formulae. This paper argues that the most commonly quoted formulae do not adequately represent the practical situation and are likely to introduce bias into the system as well as tending to make estimating more difficult. A modified ranking index derived from a decision tree type of analysis is suggested as a more useful estimate of the worth of a project, but it is concluded that any such formulae will always have serious limitations because they must inevitably ignore important characteristics of the research process.