Objective : Many studies have investigated paraspinal muscle changes after posterior lumbar surgery, including lumbar fusion. However, no study has been performed to investigate back muscle changes after pedicle based dynamic stabilization in patients with degenerative lumbar spinal diseases. In this study, the authors compared back muscle cross sectional area (MCSA) changes after non-fusion pedicle based dynamic stabilization. Methods : Thirty-two consecutive patients who underwent non-fusion pedicle based dynamic stabilization (PDS) at the L4-L5 level between February 2005 and January 2008 were included in this retrospective study. In addition, 11 patients who underwent traditional lumbar fusion (LF) during the same period were enrolled for comparative purposes. Preoperative and postoperative MCSAs of the paraspinal (multifidus+longissimus), psoas, and multifidus muscles were measured using computed tomographic axial sections taken at the L4 lower vertebral body level, which best visualize the paraspinal and psoas muscles. Measurements were made preoperatively and at more than 6 months after surgery. Results : Overall, back muscles showed decreases in MCSAs in the PDS and LF groups, and the multifidus was most affected in both groups, but more so in the LF group. The PDS group showed better back muscle preservation than the LF group for all measured muscles. The multifidus MCSA was significantly more preserved when the PDS-paraspinal-Wiltse approach was used. Conclusion : Pedicle based dynamic stabilization shows better preservation of paraspinal muscles than posterior lumbar fusion. Furthermore, the minimally invasive paraspinal Wiltse approach was found to preserve multifidus muscles better than the conventional posterior midline approach in PDS group.