Study design: Retrospective study. Objectives: The traditional PLIF is routinely utilized in severe lumbar spinal stenosis to relief the nerve compression. Nevertheless, the removal of posterior tension-band structure and the denervation and atrophy of the paraspinal muscle affect the clinical efficacy. Therefore, unilateral modified PLIF combined with contralateral fenestration was performed to overcome above-mentioned drawbacks. Methods: 32 modified PLIF and 33 traditional PLIF cases were retrospectively included. Operation time, length of stay (LOS) and blood loss were recorded. VAS of low back pain and leg pain, ODI and Sf-36 score including physical function and body pain were assessed. Fusion rate, lumbar lordosis (LL), intervertebral angle (IVA) and intervertebral height index (IHI) were evaluated radiologically. Results: Modified group possessed less blood loss, shorter operation time and less LOS. Compared with traditional group, the VAS of back pain was lower at 6 months postoperatively (P < .05) and the ODI score was lower at 3 months postoperatively (P < .05) in modified group. Modified group exhibited better physical function 3 months postoperatively and lower body pain 6 months postoperatively in Sf-36 score (P < .05). No statistic difference in LL, IVA, IHI and fusion rate were observed between both groups. Conclusions: Our modified PLIF combining with contralateral fenestration procedure exhibited particular advantages in comparison to traditional PLIF. The preservation of posterior tension-band structure facilitates to less low back pain, low complication rate and early functional recovery.