Although water jet technology has been considered as a feasible neuroendoscopic dissection methodology because of its ability to perform selective tissue dissection without thermal damage, problems associated with continuous use of water and the ensuing fountain-effect-with catapulting of the tissue-could make water jets unsuitable for endoscopic use, in terms of safety and ease of handling. Therefore, the authors experimented with minimization of water usage during the application of a pulsed holmium:yttrium-aluminum-garnet (Ho:YAG) laser-induced liquid jet (LILJ), while assuring the dissection quality and the controllability of a conventional water jet dissection device. We have developed the LILJ generator for use as a rigid neuroendoscope, discerned its mechanical behavior, and evaluated its dissection ability using the cadaveric rabbit ventricular wall. The LILJ generator is incorporated into the tip of a stainless steel tube (length: 22 cm; internal diameter: 1.0 mm; external diameter: 1.4 mm), so that the device can be inserted into a commercial, rigid neuroendoscope. Briefly, the LILJ is generated by irradiating an internally supplied water column within the stainless steel tube using the pulsed Ho:YAG laser (wave length: 2.1 microm, pulse duration time: 350 microseconds) and is then ejected through the metal nozzle (internal diameter: 100 microm). The Ho:YAG laser pulse energy is conveyed through optical quartz fiber (core diameter: 400 microm), while cold water (5 degrees C) is internally supplied at a rate of 40 ml/hour. The relationship between laser energy (range: 40-433 mJ/pulse), standoff distance (defined as the distance between the tip of the optical fiber and the nozzle end; range: 10-30 mm), and the velocity, shape, pressure, and average volume of the ejected jet were analyzed by means of high-speed camera, PVDF needle hydrophone, and digital scale. The quality of the dissection plane, the preservation of blood vessels, and the penetration depth were evaluated using five fresh cadaveric rabbit ventricular walls, under neuroendoscopic vision. Jet velocity (7.0-19.6 m/second) and pressure (0.07-0.28 MPa) could be controlled by varying the laser energy, which determined the penetration depth in the cadaveric rabbit ventricular wall (0.07-1.30 mm/shot). The latter could be cut into desirable shapes-without thermal effects-under clear neuroendoscopic vision. The average volume of a single ejected jet could be confined to 0.42-1.52 microl/shot, and there was no accompanying generation of shock waves. Histological specimens revealed a sharp dissection plane and demonstrated that blood vessels of diameter over 100 microm could be preserved, without thermal damage. The present pulsed LILJ system holds promise as a safe and reliable dissection device for deployment in a rigid neuroendoscope.