Objective Neuroendoscopic surgery allows minimally invasive surgery, but lacks effective methods to control bleeding. Water jet dissection with continuous flow has been used in liver and kidney surgery since the 1980s, and is effective for tissue manipulation with vascular preservation, but involves some potential risks, such as elevation of intracranial pressure during application in the ventricles. The authors previously reported the efficacy of the actuator-driven pulsed water jet device (ADPJ) to dissect soft tissue with vascular preservation in microscopic neurosurgery. This feasibility study investigated the use of the ADPJ to reduce the amount of water usage, leading to more safety with sustained efficacy. Methods A small-diameter pulsed water jet device was developed for use with the flexible neuroendoscope. To identify the optimal conditions for the water jet, the flow rate, water pressure, and distance between the nozzle and target were analyzed in an in vitro study by using a gelatin brain phantom. A ventricle model was used to monitor the internal pressure and temperature. For ex vivo experiments the porcine brain was harvested and ventricle walls were exposed, and subsequently immersed into physiological saline. For in vivo experiments the cortex was microsurgically resected to make the small cortico-ventricle window, and then the endoscope was introduced to dissect ventricle walls. Results In the in vitro experiments, water pressure was approximately 6.5 bar at 0.5 mm from the ADPJ nozzle and was maintained at 1 mm, but dropped rapidly toward 50% at 2 mm, and became 10% at 3.5 mm. The ADPJ required less water to achieve the same dissection depth compared with the continuous-flow water jet. With the ventricle model, the internal pressure and temperature were well controlled at the baseline, with open water drainage. These results indicated that the ADPJ can be safely applied within the ventricles. The ADPJ was introduced into a flexible endoscope and the ventricle walls were dissected in both the ex vivo and in vivo conditions. The ventricle wall was dissected without obscuring the view, and the vascular structures were anatomically preserved under direct application. Histological examination revealed that both the vessels on the ventricle wall and the fine vessels in the parenchyma were preserved. Conclusions The ADPJ can safely and effectively dissect the ventricle wall, with vascular preservation in immersed conditions. To achieve the optimal result of tissue dissection with minimal surgical risk, the ADPJ is a potential device for neuroendoscopic surgery of the ventricles.