Objective: To evaluate the pathological characteristics of endoscopic submucosal dissection (ESD) specimens for early gastric cancer and precancerous lesions, accumulating experience for clinical management and pathological analysis. Methods: A total of 411 cases of early gastric cancer or precancerous lesions underwent ESD. According to the Japanese guidelines for ESD treatment of early gastric cancer and classification of gastric carcinoma, the clinicopathological data, pathologic evaluation, concordance rate of pathological diagnosis between preoperative endoscopic forceps biopsies and their ESD specimens (in 400 cases), as well as the risk factors of non-curative resection of early gastric cancer, were analyzed retrospectively. Results: 23.4% (96/411) of the 411 cases were adenoma/low-grade dysplasia and 76.6% (315/411) were early gastric cancer. The latter included 28.0% (115/411) non-invasive carcinoma/high-grade dysplasia and 48.7% (200/411) invasive carcinoma. The concordance rate of pathological diagnosis between endoscopic forceps biopsies and ESD specimens was 66.0% (264/400), correlating with pathological diagnosis and lesion location (P < 0.01). The rate of upgraded diagnosis and downgraded diagnosis after ESD was 29.8% (119/400) and 4.2% (17/400), respectively. Among the 315 cases of early gastric cancer, there were 277 cases (87.9%) of differentiated type and 38 cases (12.1%) of undifferentiated type. In the study, 262 cases (83.2%) met with absolute indication, while 53 cases (16.8%) met relative indication. En bloc and curative resection rates were 98.1% and 82.9%, respectively. Risk factors for non-curative resection included a long diameter >20 mm (OR=3.631, 95%CI: 1.170-11.270, P=0.026), tumor infiltration into submucosa (OR=69.761, 95%CI: 21.033-231.376, P < 0.001)and undifferentiated tumor histology (OR=16.950, 95%CI: 4.585-62.664, P < 0.001). Conclusion: Several subjective and objective factors, such as the limitations of biopsy samples, the characteristics and distribution of the lesions, different pathological understanding, and the endoscopic sampling and observation, can lead to the differences between the preoperative and postoperative pathological diagnosis of ESD. In particular, the pathological upgrade of postoperative diagnosis was more significant and should receive more attention by endoscopists and pathologists. The curative resection rate of early gastric cancer in ESD was high. Non-curative resection was related to the long diameter, the depth of tumor invasion and histological classification. ESD can also be performed in undifferentiated early gastric cancer if meeting the indication criteria. The comprehensive and standardized pathological analysis of ESD specimens is clinically important to evaluate the curative effect of ESD operation and patient outcomes.