Background: Needle insertion for spinal anesthesia using the Taylor approach is challenging as the L5-S1 space is difficult to locate from the surface anatomy. In this study, we suggest the use of three-dimensional (3D) pelvis computed tomography (CT) to assist anesthesiologists in locating the needle insertion point. By comparing the success rate of 3D pelvis CT-assisted Taylor approach to that of other approaches in the existing literatures, we suggest this technique as an alternative method for subarachnoid block in the L5-S1 space. Methods: In this retrospective observational study, we reviewed the records of hip arthroplasty using the 3D pelvis CT-assisted Taylor approach. An imaginary guidance line was created from the intersection point of the midline and intercristal line on the posterior view of the 3D pelvis CT to the ideal skin insertion point for Taylor approach. The primary outcome was the success rate. The secondary outcomes included the angle between the intercristal line and the guidance line, and the length of the guidance line and the distance between the ideal needle insertion point and the L5-S1 space. Results: We reviewed the records of the 276 patients who underwent hip arthroplasty using 3D CT-assisted Taylor approach. In this cohort, the 3D CT-assisted Taylor approach in L5-S1 subarachnoid block failed in only 25/276 patients. The success rate of 3D CT-assisted Taylor approach was 90.9%. Conclusions A 3D pelvis CT-assisted Taylor approach of spinal anesthesia can be an alternative method for subarachnoid block in the L5-S1 space with an acceptable success rate.