As an illustration of the analytic approach to evaluating development projects and sectoral development programs, four diverse agricultural development strategies in Peru are analyzed in terms of net increase in farm output. The widely varying payoffs observed in the four projects reveal that too little attention was given to the problems of uncertainty, sensitivity analysis, and time discounting in planning the development strategy for each project. The uncertainty inherent in estimating cost and output for projects requires that proposals include alternative levels of costs and benefits. A comparative evaluation of the four projects suggests that intensive agricultural development activities involving large-scale capital investment should be avoided and that resources should be directed toward increasing the operating efficiency of currently established farming enterprises. 69 pp.