Farmers' attitude towards risk associated with the availability of energy inputs will influence their investment behavior and demand for energy inputs. In order to analyze policies that reduce the risk in energy availability, some modifications in methods are required. This study, using a mean-variance framework, demonstrates how cross-sectional data and time-series data on crop yields and prices can be used to analyze agricultural energy policies under uncertainty in a developing country context. It is argued that the farmers risk attitudes, their crop allocation behavior, changes in the demand for energy inputs and the stochastic relationship between various forms of energy inputs can be explained by the various energy constraints faced by them. Furthermore, using the same methods the impact of policies which affect these constraints, can be analyzed.