Agricultural development efforts quite frequently involve programs designed to extend knowledge to farm people through various kinds of extension programs. Yet, the basic methodology for evaluating such programs is not very well developed. The typical approach is to evaluate the impact of the extension service on various measures of physical productivity, or on other characteristics of technical change. This may involve the calculation of indices of partia1 and total productivity, andlor the collection of data on rates of adoption of new modern inputs or the use of new farm practices. One objective of the present paper is to suggest the application of a methodology which appears to have promise as a means of evaluating the impact of extension programs on a broader base. In addition the results of applying the methodology in a limited evaluation of the extension program of (ACAR), the extension agency for the State of Minas Gerais in Brazil, are also presented, as is a discussion of the implications of the results obtained.1984