Background: Sarcopenia represents a highly prevalent and prognostically detrimental complication in patients with bone tumors. Its progressive and dynamic nature hinders early intervention and amplifies clinical care burdens, highlighting an unmet need for tailored risk stratification tools. Objective: To determine the prevalence and independent risk factors of sarcopenia in patients with bone tumors, and to develop, internally validate, and comparatively evaluate a clinically practical nomogram for individualized sarcopenia risk prediction. Design: A prospective, single-center study was conducted. A total of 400 inpatients with bone tumors were consecutively enrolled between August 2023 and February 2025 for model development. An independent cohort of 140 patients admitted from March to June 2025 was used for head-to-head validation of the nomogram against five established sarcopenia screening tools. Setting and Participants: This study was performed at a tertiary cancer hospital, enrolling 540 patients with bone tumors using convenience sampling. Methods: (1) Model development and internal validation: Eligible patients were randomly assigned at a 7:3 ratio to a training cohort (n=280) and an internal validation cohort (n=120). Data encompassed demographics, Tampa Scale of Kinesiophobia-11 (TSK-11), Fried Frailty Phenotype, Athens Insomnia Scale (AIS), and routine laboratory parameters. Sarcopenia was defined per the 2019 Asian Working Group for Sarcopenia (AWGS) criteria. Predictors were selected using Lasso regression, and the nomogram was constructed using multivariable logistic regression. Model performance was assessed via discrimination (ROC–AUC), calibration (calibration curves and Hosmer–Lemeshow test), and clinical utility (decision curve analysis, DCA). (2) Comparative validation: The nomogram was directly compared with five brief screening tools: SARC-F, SARC-Calf, SARC-F+EBM, MSRA-7, and MSRA-5 in 140 patients. Results: Six independent predictors were integrated into the nomogram: regular exercise, body mass index (BMI), serum phosphorus, frailty, kinesiophobia, and sleep disturbance (all P<0.05). The nomogram yielded excellent discrimination, with an AUC of 0.935 (95% CI: 0.836–0.950) in the training cohort and 0.902 (95% CI: 0.808–0.910) in the internal validation cohort. The Hosmer–Lemeshow test confirmed good calibration (χ²=3.684, P=0.884). In head-to-head testing, the nomogram achieved the highest AUC (0.926), sensitivity (93.55%), negative predictive value (91.49%), Youden index (0.816), and Kappa coefficient (0.724) among all tools. Conclusions: This novel nomogram demonstrates robust predictive performance and clinical reliability for identifying sarcopenia in patients with bone tumors. It enables early, individualized risk assessment and supports timely targeted interventions to optimize clinical outcomes.