The objective of this study is to evaluate the robust regression method when detrending the crop yield data. Using a Monte Carlo simulation method, the performance of the proposed Time-Varying Beta method is compared with the previous study of OLS, M-estimator and MM-estimator in an application of crop yield modeling. We analyze the properties of these estimators for outlier-contaminated data in both symmetric and skewed distribution case. The application of these estimation methods is illustrated in an agricultural insurance analysis. The consequence of obtaining more accurate detrending method will offer the potential to improve the accuracy of models used in rating crop insurance contracts.