Purpose: There is controversy over the treatment for low grade dysplasia, while resection is recommended for high grade dysplasia. But the concordance of the grade of dysplasia between pre- and post-resection is low because of sampling errors with endoscopic biopsy. We attempted to establish a clearer direction for the treatment of dysplasia by clarifying the discrepancy between the pre- and postresection diagnoses. Materials and Methods: We performed a retrospective review of 126 patients who had undergone resection with the diagnosis of dysplasia on biopsy at Bundang CHA Hospital from 1999 to 2009. Results: Seventy patients were diagnosed with low grade dysplasia and 56 patients were diagnosed with high grade dysplasia. Among the 33 patients who received gastrectomy with lymph node dissec-tion, 30 patients were revealed to have invasive cancers and 4 patients showed lymph node metastasis. Discordance between the diagnoses from biopsy and resection occurred in 55 patients (44%). There was no correlation on the comparative analysis between the size, location or gross type of lesion and the grade of dysplasia. Conclusions: The rate of discordance between the diagnoses of endoscopic biopsy and the post resection pathologic report was as high as 44%. Endoscopic mucosal resection was not sufficient for some patients who were diagnosed with dysplasia on biopsy due to the presence of lymph node metastasis. It is necessary to be prudent when determining the follow-up and treatment based solely on the result of the biopsy.