Objective:We aimed to compare the degree of multifidus atrophy on lumbar spinal magnetic resonance imaging (MRI) between patients with degenerative lumbar spondylolisthesis (DLS) and controls. Methods:We retrospectively analysed the medical records of 213 DLS patients and recorded their baseline characteristics, including sex, age, height, weight, body mass index (weight/height2), slipped segment and presence of osteoporosis, hypertension and diabetes. The Meyerding classification system was used to categorise the patients into L3–5 spondylolisthesis groups. The L3 spondylolisthesis group consisted of 27 patients (24 and 3 with grades I and II, respectively; 6 males and 21 females) aged 43–83 years. The L4 spondylolisthesis group consisted of 140 patients (126 and 14 with grades I and II, respectively; 35 males and 105 females) aged 42–88 years. The L5 spondylolisthesis group consisted of 46 patients (40 and 6 with grades I and II, respectively; 15 males and 31 females) aged 37–91 years. In addition, we enrolled 20 young healthy volunteers aged 20–34 years as controls (12 males and 8 females). The lean and gross cross-sectional areas (LCSA and GCSA, respectively) of multifidus on both sides of L3/4, L4/5, and L5/S1 in the three groups were measured on lumbar spinal MRI. The LCSA/GCSA ratio indicated the degree of multifidus atrophy, with a larger value correlating with less multifidus atrophy. Quantitative variables were compared between the groups using independent samples t-test. Categorical data were analysed using chi-square test. Analysis of variance was performed to determine the differences in the degree of multifidus atrophy between the groups. Pearson’s linear analysis was performed to identify the correlation between slipped and non-slipped segments.Results: There was no statistically significant difference between the DLS groups in terms of baseline characteristics (p > 0.05 for all). The LCSA/GCSA ratio was significantly different between the DLS and control groups at the three levels (p < 0.001). There were significant differences in the LCSA/GCSA ratio between different levels of the same slipped segment, and between different slipped segments at the same level (p < 0.001). No statistically significant difference was observed in the LCSA/GCSA ratio between the three DLS groups with different degrees of slippage at the same segment (p > 0.05). A slight positive correlation was observed between the slipped and non-slipped segments in the L3 spondylolisthesis group (r = 0.357, p > 0.05; r = 0.078, p > 0.05, respectively). There was a significant positive correlation between the slipped and non-slipped segments in the L4 spondylolisthesis group (r = 0.686, p < 0.001; r = 0.744, p < 0.001), and a moderate positive correlation between the slipped and non-slipped segments in the L5 spondylolisthesis group (r = 0.482, p < 0.001; r = 0.448, p < 0.05).Conclusions: DLS patients demonstrated multifidus atrophy that is most prominent at the level of the slipped segment. There was a positive correlation in the degree of multifidus atrophy between the slipped and non-slipped segments, suggesting that multifidus atrophy at the level of the non-slipped segment accompanies that at the slipped segment. The multifidus atrophy was not worse in the grade II compared to the grade I spondylolisthesis group. However, the causality between DLS and multifidus atrophy remains unclear.