This paper estimates per capita demand for diesel by agricultural sector in Iran using Auto-Regressive Distributed Lags (ARDL) method over 1973-2012. The findings indicate that the demand for diesel is relatively inelastic in response to changes in its price, electricity acts as substitute input for diesel, diesel is a necessary input in agriculture, and mechanization index, despite having expected positive sign, fails to explain variations in diesel use. In addition, there is a long-run relationship among variables. According to Error Correction Model, the speed of adjustment in short-run distortions towards long-run equilibrium is moderate. The findings show that government policy on cutting energy subsidies is not enough for reducing diesel consumption in Iranian agricultural sector. Thus, non-price measures such as innovations in inter-fuel substitution technologies and applying efficient machineries are recommended to manage energy uses in agriculture.