The objective of this paper is to develop the conceptual linkage between the production function and economic surplus approaches used to evaluate the returns to agricultural research. First it is established that, for the Cobb-Douglas functional form generally used, an increase in research expenditure will result in a pivotal divergent shift in the supply function. It is then shown that the measure of this change in economic surplus is approximately the same as the value marginal product. Finally it is concluded that if the other possible types of supply shifts, that is, proportional divergent, parallel or convergent, are appropriate, then the production function approach will under-estimate the benefits and therefore rate of return to research expenditure.