With the growing interest in human capital theory, analyzing the impact ofrneducation on efficiency has been given special emphasis. this study firstrntestes the celebrated hypothesis that 'traditional farmers are efficient butrnpoor'. the result of the stochastic profit frontier function does not supportrnthis hypothesis in the Ethiopian case. the result shows that there arernconsiderable amount of deviations from the optimal profit efficiency level. itrnspecifically shows that the mean level of profit efficiency in the sampledrnfanners is 54.0 percent. this implies that there is 46.0 percent profitrninefficiency in the sampled farmers. next it is tried to test the hypothesis ofrnequal allocative and technical efficiency of educated and illiterate farmers byrnusing the modified Y-L profit function model under various linearrnrestrictions. the results reveal that educated farmers are relatively andrnabsolutely more efficient than illiterate fanners. this implies that at thernexisting level of factor endowments and technology there is a potential tornincrease agricultural output by expanding education and consequently byrnmaking illiterate farmers to operate closer to the efficiency level achieved byrntheir educated neighbors. It is also shown that education increases not onlyrnthe efficiency of farmers but also the probability of farmers to adoptrnimproved inputs such as fertilizer. the multinomial probit analysis alsornshows that education and environment variables are substitutes in modemrnareas and complementary in traditional environments. this suggests that expansion of schools and increasing enrollment rates in rural areas havernhigher payoff than in modem areas at least in increasing the probability ofrnfarmers to adopt fertilizer input.