A standard model of behavior under uncertainty is used to suggest price risk variables for use in a positive supply study. The suggested variables are intuitively appealing and empirically tested on Pinto bean data. Linearity is assumed and O.L.S. used. The empirical results show that the risk variables greatly improve the statistical fit of the supply equation, are quantitatively important and that a substantial bias occurs if they are neglected. Policy initiatives to reduce Pinto bean price fluctuations need to consider the risk reducing effects on the supply response.