Abstract The agricultural sector is undoubtedly the core of economic development. Although the sector is subject to intense protectionism, the number of lawsuits for dumping agricultural products filed has not increased noticeably. Blonigen (2003) discovered that in all lawsuits filed by WTO members, the successful cases of antidumping agricultural products only accounted for 6 %. This study reveals that difficulties in identifying damage standards and in collecting tangible information are the chief reasons explaining the low percentage of filed lawsuits of antidumping agricultural products. Hence we need a more objective, accurate and straightaway measurement for industrial damage identification. In our study, we applied COMPAS model for analysis, and we found that a probabilistic model that takes into account the uncertainty surrounding economic parameter reduces the risk of misjudging the effect of dumping on the domestic industry. Based on the statistic data by Prusa (2001), most of the filed lawsuits of antidumping agricultural products are in fruits, sugar and meat. We chose coconut and persimmon products as our research targets. Simulation results show that dumping will cause injury to coconut and persimmon industries. And limiting the analysis to the 8 scenarios of COMPAS model might tend to overstate the overall effect on the domestic industry in terms of quantity of output.