The sarcopenia index (SI), calculated as [(serum creatinine/serum cystatin C) × 100], maybe a simpler alternative for measuring muscle mass than computed tomography (CT) and bioelectrical impedance analysis (BIA). We enrolled 112 patients with head and neck cancers (HNC). The correlation of the SI with muscle surface area measured by CT (CTMSA, n = 82) and muscle mass by BIA (BIA-MM, n = 41) was tested. Cutoff values were set for SI, CTMSA, and BIA-MM. Overall survival (OS) was compared between the high- and low-SI/CTMSA/BIA-MM groups. The SI was correlated with CTMSA (r = 0.43) and BIA-MM (r = 0.52). The optimal cutoff values of SI, CTMSA, and BIA-MM were 76.1 (area under the curve [AUC] = 0.67), 129.2 (AUC = 0.59), and 46.1 (AUC = 0.62), respectively. OS was significantly lower in the low-SI group (78% at 1 year and 69% at 2 years) than in the high-SI group (94% at 1 year and 86% at 2 years; p = 0.006). There was no significant difference in OS between the low-and high-CTMSA and -BIA-MM groups. The SI, which only requires a blood sample, is a useful marker of muscle mass that correlates with short-term prognosis in patients with HNC.