Abstract:[Objective] To verify the effect of "shoulder-to-shoulder" anatomical marker localization in restoring lower limb length dur-ing total hip arthroplasty (THA). [Methods] "Shoulder-to-shoulder" THA was performed in 4 cadavers based on insertion obturator exter-nus as the “shoulder” of the femur for osteotomy. The length of lower limb, tip-shoulder distance, osteotomy distance, longitudinal andhorizontal deviation of center of rotation (COR) of femoral head, acetabular abduction angle and acetabular anteversion angle were mea-sured, and verified with the artificial intelligence hip (AIHIP). [Results] All the 4 cadavers got equal length of bilateral lower limbs after op-eration. There were no statistically significant differences in tip-shoulder distance, osteotomy distance, longitudinal and horizontal devia-tions of femoral head COR between the postoperative imaging measurement and the preoperative AIHIP planning (P＞0.05). However, theacetabular abduction angle measured by postoperative image was significantly lower than that planned by AIHIP [(32.5±3.1)° vs (41.3±2.5)°, P=0.005], while the acetabular anteversion measured postoperatively were significantly greater than that planned by AIHIP [(25.3±9.4)° vs (18.8±2.5)°, P=0.030]. In addition, The tip-shoulder distance measured intraoperatively was significantly greater than that mea-sured by postoperative imaging [(26.3±2.6) mm vs (15.5±1.5) mm, P＜0.001], and preoperative AIHIP planning [(26.3±2.6) mm vs (17.0±2.2) mm, P＜0.001]. [Conclusion] The validation of AIHIP after cadaveric surgery shows that the "shoulder-to-shoulder" as anatomic mark-er for femoral stem placement does ensure the accurate placement and effectively restore the lower limb length, although some parametersmeasured at different time points are different.