Scientific and technical information services cost the federal government close to a billion dollars a year and cost further large sums to other sectors of the economy. Unless the relationship of benefit to cost for these services is quantitatively understood, there are dangers of enormous waste in some areas or of the passing up of enormous net benefits in others. Very little has been done toward the assignment of a monetary value to these benefits. It should be possible to develop rough but useful quantitative measures for the benefits delivered by many kinds of information services by combining theoretical analysis with suitably gathered empirical data on the response of the user community to these or related services. The two most useful sources of information on quantified judgements of value by users are market response--what users, as individuals or through their organizations, are willing to pay for an information product or service--and investment of time by individual users. To obtain these value measures, extensive data on the markets for information services and on the habits of their users will be needed, as well as sophisticated theoretical analysis of markets and of the interaction of different modes of information transfer. (Author/PF)