Price in agricultural supply equations is usually the expected price. In general, models of agricultural supply response assume that their representation of producer expectations is correct. If this assumption is wrong, the supply response parameter will have embodied within it an estimate of the expected price distortion, biasing the estimated parameter. Furthermore, no effort has been made in existing models to allow for heterogeneous price expectations. In almost every supply model, one price estimate is used to represent the price expectations of hundreds or thousands of producers, which could masks a wide range behavior among heterogenous producers. Rather than assuming the model’s representation of producer expectations is correct, the goal of this paper is to use revealed producer behavior to assist in the estimation of producer’s expected prices. To explicitly address producers’ formation of the expected price, this study jointly estimates the supply equation and the price expectation equation. The empirical application is a pooled time-series cross-section data analysis of U.S. wheat and corn supply data at the county level.