Background and Aims: Endoscopic submucosal dissection (ESD) is considered technically difficult and challenging using a conventional flexible endoscope, mainly due to the lack of proper countertraction to expose the submucosal dissection plane. This study aimed to evaluate the feasibility of a traction method using a dexterous robotic arm in ex vivo gastric ESD. Methods: ESD was performed in a total of 45 procedures using a portable endoscopic tool handler (PETH) (n = 30) and using the conventional method (n = 15) at various locations in the stomach. For each procedure, the performance data were recorded, including the total procedure time (minutes), incision time (minutes), dissection speed (mm(2)/minute), and blind dissection rate (%), to enable a comparison of the 2 ESD methods. Results: The total procedure time was significantly shorter with PETH-ESD than in conventional ESD (23 vs 36 minutes, P = .011). This result is mainly attributed to the dissection speed, which was significantly faster, by more than 2.5 times, using the PETH (122.3 +/- 76.5 vs 47.5 +/- 26.9 mm(2)/minute, P <.001). The blind dissection rate was greatly decreased in PETH-ESD (0 vs 20%, P <.001). There was no significant difference in the incision time (6.1 +/- 5.0 vs 5.5 +/- 2.9 min, P = .612). Conclusions: The countertraction method using the PETH significantly improved the dissection speed and reduced blind dissection by enhancing direct visualization of the submucosal plane. With the advantages of multidirectional traction, fine tension control, and regrasping, this new device is expected to improve the performance of ESD and further facilitate advanced endoscopic procedures.