The green revolution in developing countries magnified concern about the efficient allocation of agricultural research resources and the distributional consequences of alternative research resource allocation and technology design strategies. These concerns are being increasingly reflected in the planning, management and research activities of the international agricultural research centres. In this paper a description is given of how economists at one centre contributed their expertise to the ex anfe analysis of some key issues in this complex milieu, such as the determination of research goals and priorities, the small-large farm dichotomy or nexus, income distribution and employment effects, human nutritional considerations and farmer risk attitudes.Agricultural research in developing countries has had a high payoff, both in terms of the return on the investments required, and in terms of equity. The 23 studies of agricultural research productivity in developing countries reviewed by Ruttan (1980) had an average annual rate of return on investment of 55 per cent. In the studies, research on food crops, fivestock and commercial crops such as cotton and rubber was covered. These high rates of return to agricultural research in developing countries suggest that the levels of research investment remain well below what they should be to exploit fully the opportunities for increased agricultural production and enhancement of economic development and human welfare.Economists have played a key role in assessing the impact of past investments in agricultural research, which is amply displayed in Ruttan's (1980) analysis and in Arndt, Dalrymple and Ruttan (1977). Their role in the ex ante analysis of research resource allocation strategies is growing and seems not yet to have reached its full potential.With the increasing trend toward inclusion of economists in agricultural research institutions and agricultural ministries in developing countries, their scope to contribute constructively to the design of technology options and policies is enhanced. However, ex ante evaluation is fraught with many uncertainties, not the least for economists. Considerable subjectivity and intuition is still required and, in these circumstances, use of formal planning models of the type described in Fishel (1 971) often only serve to cloud the decision-making environment with an air of certainty and objectivity which can be inappropriate (Arnon 1975, p. 51). Many of the techniques of project appraisal were designed for industry where costs and gains are generally private. In * I am grateful to Tom Walker, Don Winkelmann and the three Journal referees for constructive comments on an earlier draft. The views expressed in this paper do not necessarily reflect those of ACIAR.