Objective: To assess the feasibility of HEAD-US scale in the clinical application of hemophilic arthropathy (HA) and propose an optimized ultrasound scoring system. Methods: From July 2015 to August 2017, 1 035 joints ultrasonographic examinations were performed in 91 patients. Melchiorre, HEAD-US (Hemophilic Early Arthropathy Detection with UltraSound) and HEAD-US-C (HEAD-US in China) scale scores were used respectively to analyze the results. The correlations between three ultrasound scales and Hemophilia Joint Health Scores (HJHS) were evaluated. The sensitivity differences of the above Ultrasonic scoring systems in evaluation of HA were compared. Results: All the 91 patients were male, with median age of 16 (4-55) years old, including 86 cases of hemophilia A and 5 cases hemophilia B. The median (P₂₅, P₇₅) of Melchiorre, HEAD-US and HEAD-US-C scores of 1 035 joints were 2(0,6), 1(0,5) and 2(0,6), respectively, and the correlation coefficients compared with HJHS was 0.747, 0.762 and 0.765 respectively, with statistical significance (P<0.001). The positive rates of Melchiorre, HEAD-US-C and HEAD-US scale score were 63.0% (95%CI 59.7%-65.9%), 59.5% (95%CI 56.5%-62.4%) and 56.6% (95%CI 53.6%-59.6%) respectively, and the difference was statistically significant (P<0.001). Even for 336 cases of asymptomatic joints, the positive rates of Melchiorre, HEAD-US-C and HEAD-US scale score were 25.0% (95%CI 20.6%-29.6%), 17.0% (95%CI 12.6%-21.1%) and 11.9% (95%CI 8.4%-15.7%) respectively, and the difference was statistically significant (P<0.001). There were significant changes (P<0.05) in the ultrasonographic score of HA before and after onset of hemorrhage in 107 joints of 40 patients. The difference in variation amplitude of HEAD-US-C scores and HEAD-US scores before and after joint bleeding was statistically significant (P<0.001). Conclusion Compared with Melchiorre, there were similar good correlations between HEAD-US, HEAD-US-C and HJHS. HEAD-US ultrasound scoring system is quick, convenient and simple to use. The optimized HEAD-US-C scale score is more sensitive than HEAD-US, especially for patients with HA who have subclinical state, which make up for insufficiency of sensitivity in HEAD-US scoring system