Purpose: To determine the prognostic role of muscle area and muscle radiation attenuation in the erector spinae (ES) and multifidus (MF) muscles in patients undergoing gastrectomy. Methods: Patients with stage I-III gastric cancer undergoing gastrectomy were retrospectively enrolled in this study. Clinicopathologic characteristics were collected and analyzed. Both paraspinal muscle index of ES/MF muscles (PMIEM) and paraspinal muscle radiation attenuation in the same muscles (PMRAEM) were analyzed at the 3rd lumbar level using axial computed tomographic images. Cox regression analysis was applied to estimate overall survival (OS) and disease-free survival (DFS). Results: There was only a weak correlation between PMIEM and PMRAEM (r=0.28). Multivariate Cox regression revealed that PMRAEM, but not PMIEM, was an important determinant of survival. PMRAEM along with age, tumor-node-metastasis (TNM) stage, perineural invasion, and serum albumin level were significant determinants of both OS and DFS that constituted Model 1. Harrell’s concordance index and integrated area under receiver operating characteristic curve were greater for Model 1 than for Model 2 (consisting of the same covariates as Model 1 except PMRAEM) or Model 3 (consisting of only TNM stage). Conclusion: PMRAEM, but not PMIEM, was an important determinant of survival. Because there was only a weak correlation between PMIEM and PMRAEM in this study, it was presumed that they were mutually exclusive. Model 1 consisting of age, TNM stage, perineural invasion, serum albumin level, and PMRAEM was greater than nested models (i.e., Model 2 or Model 3) in predicting survival outcomes.