Abstract Background Context An overload of the paravertebral muscles after surgical intervention is suggested to be the major cause of postoperative pain. In cross-sectional area analyses, increased atrophy of the multifidus muscle (MF) after conventional open versus minimally invasive posterior lumbar interbody fusion (PLIF) has been described. The three-dimensional characteristics of the paravertebral muscles and separate evaluation of the longissimus muscle (LS) have not been addressed in analyses thus far. Purpose The purpose of the present study was to compare the MF and LS volume atrophy and fatty degeneration between single-level minimally invasive transforaminal lumbar interbody fusion (miTLIF) and conventional midline approach-based PLIF (coPLIF) of L4/L5 or L5/S1 at the index and superior adjacent segments. Design This was a prospective, randomized, controlled, non-blinded study. Patient Sample Fifty patients with single-level segment degeneration (Pfirrmann ≥III and Modic ≥3) of L4/L5 or L5/S1 not requiring decompression were randomly assigned to two groups. Outcome Measures Paraspinal lumbar residual muscle tissue volume, change in the relative fat content of MF and LS at the index and superior adjacent segments, and clinical parameters, including a visual analogue scale (VAS) for low back pain and the Oswestry Disability Questionnaire (ODI) were the outcome measures in this study. Methods Twenty-five patients were treated with miTLIF, and the remaining patients were treated with coPLIF (both with transpedicular fixation). Clinical scoring was performed preoperatively and at 1 week and 12 months postoperatively, and computed tomography was performed at the latter two follow-ups. Results The LS damage at the index segment was similar in both groups (3% greater fat content increase in the coPLIF vs. the miTLIF group, p=.032), whereas MF atrophy and degeneration were increased (p<.001) in the coPLIF group. At the adjacent segment, muscle atrophy and increased fatty infiltration (p<.05) were minimal in both muscles but were similar in both groups. Visual analogue scale and ODI scores improved (p<.001), without differences between the groups. Conclusions The muscle damage after miTLIF was inferior to that after coPLIF; spatially, however, the muscle sparing was predominantly attributed to the MF and, surprisingly, not to the LS.