Objective Endoscopic submucosal tunnel dissection (ESTD) is a new treatment for superficial upper gastrointestinal precancerous lesions and tumors. However, its safety and efficacy for treating these lesions is still uncertain. This meta‐analysis aimed to compare the efficacy and safety of ESTD with those of endoscopic submucosal dissection (ESD) in treating early superficial upper gastrointestinal precancerous lesions and tumors. Methods Electronic databases, including EMBASE, PubMed, Cochrane Library and the Web of Science, were systematically searched for articles published up to 1 April 2020. The primary outcomes were complete and en bloc resection rates of the lesions. The secondary outcomes included operation time, speed of dissection, complication and recurrence rates. The Newcastle‐Ottawa quality assessment scale was used to evaluate the quality of the studies. Results Seven studies were included in the final meta‐analysis. The complete and en bloc resection rates of early superficial esophageal lesions as well as the en bloc resection rate of early superficial gastric lesions were similar between the ESTD and ESD groups. Compared with the ESD group, the ESTD group had a higher en bloc resection rate for early superficial gastric lesions. A faster dissection speed was observed in both superficial esophageal and gastric lesions. In gastric lesions, the operation time of the ESTD group was shorter than that of the ESD group, while it did not differ for esophageal lesions. There were no obvious differences in bleeding and perforation rates between the groups. Conclusion As an endoscopic technique for treating early superficial upper gastrointestinal lesions, ESTD is as safe as and more effective than ESD, especially for gastric lesions.