Although interlaminar endoscopic lumbar discectomy (IELD) is considered to be less invasive than microscopic lumbar discectomy (MLD) in treatment of lumbar herniated nucleus pulposus, the radiologic change of multifidus muscles by each surgery has rarely been reported. The aim of the present study was to compare the quantitative and qualitative changes of multifidus muscles between two surgical approaches and to analyze the correlation between various parameters of multifidus muscles and long term surgical outcome. 21 patients who received MLD and 18 patients who received IELD in a single tertiary hospital were enrolled and their preoperative, postoperative (≤15 days), and follow-up (≥6 months) MRIs were analyzed. The cross-sectional area (CSA) and fatty degeneration rate (FD) were quantitatively estimated at the level of surgery. The correlations among CSA, FD, body mass index, follow-up visual analogue scale(VAS) and Oswestry Disability Index(ODI) were assessed. Mean intervals of postoperative MRI and follow-up MRI from surgery were 3.0±3.7 days and 14.5±10.7 months, respectively. During the follow-up period, VAS was improved from 7.1±1.3 to 2.1±1.8 in MLD and from 8.2±1.4 to 2.2±1.8 in IELD. In cases of MLD, comparing with preoperative MRI, ipsilateral CSA was significantly increased in postoperative MRI (795.6mm. 2. vs. 906.5mm. 2. , p<0.01), but it was not significantly different in follow-up MRI (795.6mm. 2. vs. 814.4mm. 2. , p=1.00). However, in case of IELD, the ipsilateral CSAs in preoperative, postoperative, and follow-up periods were 892.0 mm. 2. , 909.3 mm. 2. , and 900.3 mm. 2. , respectively. These changes were not significant over time (p=0.691). The ipsilateral FDs were not significantly changed between preoperative and follow-up periods in both MLD (21.4% vs. 20.9%, p=0.81) and IELD groups (23.5% vs. 21.8%, p=0.19). The increment of ipsilateral CSA had significant correlations with follow-up ODI (r=−0.368, p=0.02). Comparing with IELD, MLD induced more surgical trauma on multifidus muscle in postoperative period, but the muscular damage was recovered in follow-up period. IELD can minimize surgical trauma on multifidus muscle showing similar pain relief as MLD. Favorable surgical outcome in follow-up period may be related to increment of multifidus muscle volume. Figure 1. (A-C) The multifidus muscles in preoperative, postoperative, and follow-up periods, respectively, in patient with MLD. Comparing with preoperative period, the CSA of right multifidus muscle (ipsilateral side) was increased in postoperative period, but recovered in follow-up period. (D-F) The multifidus muscle in preoperative, postoperative, and follow-up periods, respectively, in patient with IELD. The CSA of left multifidus muscles (ipsilateral side) was not significantly changed over time. Comparing preoperative MRIs with follow-up MRIs, the FDs of multifidus muscles were not significantly changed regardless of surgical technique. Figure 2. The CSA was measured by marking region of interest (ROI) and FD was measured by calculating the rate of pixels beyond the threshold in ROI. All measurements were performed using ImageJ software (version 1.52a, National Institutes of Health, Bethesda, Maryland, USA). For any figures or tables, please contact the authors directly