Development of genetically modified crops is challenging the functions of the grain marketing system with many participants arguing for Identity Preservation systems prior to release of GM varieties. In this study, a stochastic optimization model was developed to determine optimal testing strategies. The model chooses the optimal testing strategy that maximizes utility (minimizes disutility) of additional system costs due to testing and rejection and allows estimation of the risk premium required for sellers to undertake a dual marketing system with GM/Non-GM segregations over the current Non-GM system. Elements of costs (testing costs, rejection costs, and risk premium) were estimated for a base model representing a grain export chain. The model includes elements of costs and risks for uncertainties within the marketing chain including risk of adventitious commingling at all stages of the marketing chain, grower truth-telling, variety declaration, and accuracy of testing technologies. Sensitivities were evaluated for effects of GM adoption, risk parameters, variety declaration, tolerance levels, and for a domestic market case.