The PROCISUR program of support to National Agricultural Research Centers (NARCs) in Brazil, Argentina, Chile, Paraguay and Bolivia entails specialized activities designed to enhance international transfer of research findings. An evaluation of the program requires that this enhancement feature be built into the evaluation model. This requires some degree of methodological extension to existing methods. In this paper it is developed such an extension and apply it to an evaluation of the PROCISUR program as it affects productivity change in wheat, maize, and soybean production in the member countries since the inception of the program in 1978. The table of contents is: Section II provides a summary of the objectives of the PROCISUR program and of the resources devoted to program activities. Section III then develops the methodological extensions required for data analysis and estimates of program impacts. Section IV reports our estimates of PROCISUR impacts. The final section draws policy conclusions. (MIBA)