Background: Increasing evidence links sarcopenia and cancer prognosis, but limited data have focused on whether and to what extent muscle radiodensity can impact cancer outcomes. This study was conducted to investigate whether skeletal muscle mass, when divided into subranges of low or high radiodensity, improves prediction of short-term survival in patients with endometrial cancer (EC). Four skeletal muscle phenotypes were proposed to assess which is the best predictor of 1-year mortality. Methods: Patients with EC who had CT images available within 30 days before treatment (n=208) were enrolled in a retrospective cohort. CT images at the third lumbar vertebra (L3) were used to assess overall skeletal muscle index (SMI), which was then divided into subranges of radiation attenuation: low- and high-radiodensity SMI. The average muscle radiation attenuation (AMA) was also assessed. SMI and AMA were categorized as below or above the median and as below or above 30 Hounsfield units (HU), respectively, to construct 4 skeletal muscle phenotypes: “high SMI + high AMA”; “low SMI + high AMA”; “high SMI + low AMA”; and “low SMI + low AMA”. One-year survival was evaluated using the Kaplan-Meier method and Cox multiple regression analysis. Results: All of the skeletal muscle parameters, except the SMI, were significantly associated with shorter 1-year survival. The skeletal muscle phenotype of “low SMI + low AMA” showed the strongest association with 1-year mortality (hazard ratio, 5.36; 95% CI, 1.70–16.51). Conclusions: The additional value of classifying the skeletal muscle into subranges of radiodensity should be explored in the future. Evaluating the impact of skeletal muscle phenotypes on cancer prognosis is promising and must be assessed in further studies.