We evaluated the biomechanical behavior of the endoscopic decompression for lumbar spondylolysis using the finite element technique. An experimentally validated, 3-dimensional, non-linear finite element model of the intact L3 - 5 segment was modified to create the L4 bilateral spondylolysis and left-sided endoscopic decompression. The model of Gill's laminectomy (conventional decompression surgery of the spondylolysis) was also created. The stress distributions in the disc and endplate regions were analyzed in response to 400 N compression and 10.6 Nm moment in clinically relevant modes. The results were compared among three models. During the flexion motion, the pressure in the L4/5 nucleus pulposus was 0.09, 0.09 and 0.16 (MPa) for spondylolysis, endoscopic decompression and Gill's procedure, respectively. The corresponding stresses in the annulus fibrosus were 0.65, 0.65 and 1.25 (MPa), respectively. The stress at the adjoining endplates showed an about 2-fold increase in the Gill's procedure compared to the other two models. The stress values for the endoscopic and spondylolysis models were of similar magnitudes. In the other motions, i. e., extension, lateral bending, or axial rotation, the results were similar among all of the models. These results indicate that the Gill's procedure may lead to an increase in intradiscal pressure (IDP) and other biomechanical parameters after the surgery during flexion, whereas the endoscopic decompression did not change the segment mechanics after the surgery, as compared to the spondylolysis alone case. In conclusion, endoscopic decompression of the spondylolysis, as a minimally invasive surgery, does not alert mechanical stability by itself.