Background: Noncompressible torso hemorrhage (NCTH) is the main cause of prehospital death due to war injury. External hemorrhage control devices (EHCDs) are effective in reducing the death risk of NCTH, but the pressurized area is too large to prevent serious organ damage. This study aims to establish the surface localization strategy of EHCDs based on anatomical features of NCTH-related arteries through human CT images to facilitate the optimal design and application of EHCDs and eventually reduce relevant injuries. Methods: Images of 200 patients aged 18-65 years who underwent abdominal contrast-enhanced CT scans were collected. Anatomical parameters such as the length and diameter of the aortic bifurcation (AB), common iliac artery (CIA), external iliac artery (EIA) and common femoral artery (CFA) were measured, and positional relationships among EHCD-targeted arteries, umbilicus, anterior superior iliac spine (ASIS) and pubic tubercle (PT) were determined. The accuracy of surface localization was verified by 3D-printed mannequins of 20 real patients. Results: The angle and diameter of the AB were 47.1±10.5° and 17.8±2.7 mm, respectively. The AB was 7.5±8.6 mm to the left of umbilicus. The left CIA and right EIA were longer than the contralateral CIA and right EIA (p=0.038 and p=0.000, respectively). The average length of CFA was 39.0±16.6 mm. The CIA was wider than the EIA and the CFA (p=0.000). The distance between the artery and surface decreased from the CIA to the CFA (p=0.000). The vertical distance between the CIA terminus and the ipsilateral AB-ASIS line was 19.6±8.2 mm, and the left and right perpendicular intersections were located at the upper 1/3 and 1/4 of the AB-ASIS line, respectively. The length ratio of EIA-ASIS to ASIS-PT was 0.6:1. The length of the surface location/actual subpoint-ASIS was significantly correlated (p≤0.002), and the vertical distance between the two points of the same artery was ≤ 5.5 mm. Conclusion: The arterial localization strategy established via anatomical investigation was consistent with the actual situation. The data are necessary for improving EHCD design, precise hemostasis and EHCD-related collateral injuries.