The estimation of technical coefficients of production is a prerequisite for the construction of an agricultural production model. The production functions, introduced by the classical school and developed extensively by the neo-classical writers, have been frequently used in deriving the technical coefficients. But the problems posed by the management factor  and simultaneous determination of variable inputs and outputs by the farm firm [5, 10) are quite serious and have generally led to biased estimates. On the other hand, the profit function [14, 19] has been developed as an alternative to the production function. Since the arguments of this function are normalized input prices and fixed inputs which are exogenously determined, the bias of simultaneous equations is avoided. However, the profit function may be difficult to estimate due to data problems. For example, what is the wage rate of agricultural labour and how is it imputed? Also, data on product pi ices are not easily accessible and some products may not be marketed at all, especially, in developing countries.