A variety of crop revenue insurance programs is now available. These programs require measurement of price risk. This article investigates the appropriateness of distributional assumptions underlying current and proposed alternative actuarial methods. Our results find that prices are best modeled using a flexible mixture of normals distribution.