Reduced muscle mass (RMM) is a phenotypic criterion for malnutrition; the appendicular skeletal muscle mass index (ASMI) and fat-free mass index (FFMI) are both applicable indicators in the global leadership initiative on malnutrition (GLIM) guideline. However, their sensitivity and prognostic effect remain unclear. Clinical data of 2,477 patients with malignant tumors were collected. Multi-frequency bioelectrical impedance analysis was used to obtain ASMI and FFMI. RMM was confirmed by ASMI (< 7.0 kg/m2 for men and < 5.7 kg/m2 for women) or FFMI (< 17 kg/m2 for men and < 15 kg/m2 for women). Propensity score match analysis and logistic regression analysis were used to evaluate the efficacy of FFMI and ASMI in diagnosing severe malnutrition and multivariate Cox regression analysis to determine the efficacy of RMM in predicting survival. In total, 546 (22.0%) and 659 (26.6%) participants were diagnosed with RMM by ASMI (RMM.ASMI group) and FFMI (RMM.FFMI group); 375 cases overlapped. Body mass index (BMI), midarm circumference, triceps skinfold thickness, and maximum calf circumference were all significantly larger in the RMM.FFMI group for both sexes (P < 0.05). A 1:1 matched dataset constructed by propensity score match contained 810 cases. RMM.FFMI was an influential factor of severe malnutrition with HR = 3.033 (95% CI 2.068-4.449, P < 0.001), and RMM.ASMI was a predictive factor of overall survival (HR = 1.318, 95% CI 1.060-1.639, P = 0.013 in the RMM.ASMI subgroup, HR = 1.315, 95% CI 1.077-1.607, P = 0.007 in the RMM.FFMI subgroup). In general, RMM indicates negative clinical outcomes; when defined by FFMI, it predicts nutritional status, and when defined by ASMI, it is related to poor survival in cancer patients.