To obtain the quantitative measurements of the muscle morphology of cervical extensors in patients with multilevel cervical spondylotic myelopathy, and determine whether the morphological parameter of each muscle correlates with the patients' demographic features, symptoms, and cervical sagittal balance. We retrospectively evaluated 100 hospitalized patients scheduled to undergo surgery for multilevel cervical spondylotic myelopathy. Demographic data, including age, sex, and body mass index, were recorded, and symptoms were evaluated using the visual analog scale (VAS), neck disability index (NDI), and modified Japanese Orthopedic Association scale scores. The cross-sectional area (CSA) of the multifidus, semispinalis capitis, semispinalis cervicis, splenius capitis and splenius cervicis were measured on magnetic resonance imaging. The CSA of the total extensor muscles, deep extensor muscles (DEM, consisting of multifidus and semispinalis capitis), and superficial extensor muscles (consisting of semispinalis cervicis, splenius capitis and splenius cervicis) were calculated. The adjusted CSA (aCSA) was calculated as the CSA of the muscle/CSA of the corresponding vertebral body. The fat infiltration ratio (FIR) of the posterior extensor muscles was assessed using a pseudocoloring technique. Sagittal parameters, including cervical lordosis (CL), C2-7 cervical sagittal vertical axis (SVA), T1-slope, mismatch between T1-slope and CL (T1S-CL), and range of motion, were measured. The measured parameters were compared between the males and the females, between the patients with higher muscle aCSA and the patients with lower muscle aCSA, and between the patients with and without sagittal balance. A Pearson correlation analysis was conducted to determine the correlations between the paraspinal muscle measurements, and the clinical and radiographic parameters. There were 67 males and 33 females in this study, and the mean age was 59.22±9.54years. Compared with females, male patients showed higher CSA and aCSA of extensor muscles. Patients with lower muscle aCSA were significantly older and had worse NDI scores, with significantly greater C2-7 SVA and T1S-CL. Patients with sagittal imbalance showed significantly lower aCSA of total extensor muscles and DEM, as well as a significantly higher FIR. Age was significantly correlated with the aCSA of each measured muscle and the FIR. The aCSA of the DEM was correlated with the NDI score, the visual analog scale score, the SVA, the T1-slope, and the T1S-CL. In patients with multilevel CSM, age and sex were demographic factors that were highly correlated with muscle morphology changes. Extensor muscles, especially DEM, play important roles in maintaining cervical sagittal balance and are associated with the severity of neck symptoms.