This paper investigates possibilities of reducing production inefficiency using same input vector. Firms mostly suffer technical inefficiency in their production. Producers are found to operate below the frontier. Hence, empirical measures of production efficiencies are necessary to determine the inefficiency level, magnitude of inefficiency reduction, and gains that could be obtained by improving performance in the sector. The study used secondary data collected from Food and Agriculture Organization (FAO) and the National Bureau of Statistics (NBS) database. The dataset covers the period 1960-2021. Variables extracted included agricultural GDP (kg), fertiliser usage (kg), agricultural labour(man-days), and number of tractors (No). The data were subjected to a unit root test for stationarity, and a stochastic frontier production/ Cost function model were applied to determine technical inefficiency of variables in the model. The result of unit root shows that series are integrated of the first order I (1). AIC criteria indicate an optimal lag length of two years, while the Unrestricted Co-integration Trace and Maximum Eigenvalue test show strong evidence of long-run relationship amongst variables. The parameters of Stochastic Frontier Production Function estimated were positively consistent with the study’s a priori expectation for such variables as fertiliser (0.2634), labour (0.3159), land (0.1846), and tractor (0. 1587). Output-oriented technical efficiency is 0.7802 (78%) and 22% is technically inefficient. A decreasing returns to scale value of (0.9226) with a scale effect of 0.9226. The sector is 58% economically efficient with cost savings of 42% and 75% allocative efficiency. This study concludes that Nigeria’s agriculture suffers from production inefficiencies and this inefficiency can be reduced by using the same input levels. Keywords: Cost function, efficiency, Return to scale, Scale effect, and Stochastic.