The use of alternative probability density functions to specify risk in farm programming models is explored and compared to a traditional specification using historical data. A method is described that compares risk efficient crop mixes using stochastic dominance techniques to examine impacts of different risk specifications on farm plans. Results indicate that a traditional method using historical farm data is as efficient for risk averse producers as two other methods of incorporating risk in farm programming models when evaluated using second degree stochastic dominance. Stochastic dominance with respect to a function further discriminates among the distributions, indicating that a density function based on the historic forecasting accuracy of the futures market results in a more risk-efficient crop mix for highly risk averse producers. Results also illustrate the need to validate alternative risk specifications perceived as improvements to traditional methods.