BACKGROUND Endoscopic submucosal dissection (ESD) has become one of the mainstays of treatment for early gastric cancer (EGC). Radical surgery is also a classical treatment method for EGC. There have been no systematic clinical studies of the curative effects and adverse events associated with ESD vs. radical surgery for EGC. This study investigated the therapeutic efficacy and safety of ESD and radical surgery for EGC. METHODS Twenty-nine patients with EGC underwent ESD, and 59 underwent radical surgery at Weihai Municipal Hospital. The pathological characteristics, postoperative outcomes, hospital course, morbidity and mortality were retrospectively compared between the two groups. RESULTS The oncological clearance was 93.1 % (27/29) in the ESD group. Postoperative delayed haemorrhage occurred in two patients. The hospital stay ranged from 10 to 23 days, and the average stay was 14.3 ± 3.7 days. The patients were followed-up for 1 to 5 years, with a mean follow-up of 26.9 ± 8.5 months. Regular endoscopic examinations showed that the wound had healed with no cancer recurrence in all of the patients. In the radical surgery group, the oncological clearance was 100 % (59/59). The hospital stay ranged from 11 to 55 days, and the average stay was 21.7 ± 9.3 days. The patients were followed-up for 1 to 3.7 years, with a mean follow-up of 22.3 ± 9.4 months. Nine patients developed complications, including acute postoperative adhesive ileus (1/59) and symptomatic residual gastritis (3/59). These complications were improved by an additional operation, drainage, gastrointestinal decompression and comprehensive therapy. CONCLUSIONS ESD achieved similar efficacy and had many advantages compared with radical surgery for the treatment of EGC.