The orientation of the natural acetabulum is useful for total hip arthroplasty (THA) planning and for researching acetabular problems. Currently the same acetabular component orientation goal is generally applied to all THA patients. Creating a patient-specific plan could improve the clinical outcome. The gold standard for surgical planning is from threedimensional (3D) computed tomography (CT) imaging. However, this adds time and cost, and exposes the patient to a substantial radiation dose. If the acetabular orientation could be reliably derived from the standard anteroposterior (AP) radiograph, preoperative planning would be more widespread, and research analyses could be applied to retrospective data, after a postoperative issue is discovered. The reduced dimensionality in 2D Xray images and its perspective distortion, however, lead to ambiguities that render an accurate assessment of the orientation parameters a difficult task. The goal of this work is to enable robust measurement of the acetabular inclination and version on 2D Xrays using computeraided techniques. Recently, methods have been proposed to reconstruct the patientspecific 3D pelvic surface and orientation and component orientation from standard 2D postoperative Xray images. We propose a reconstruction method to determine the natural acetabular orientation from a single, preoperative Xray. The approach utilizes novel articulated statistical shape and intensity models (ASSIMs) that express the variance in anatomical shape and bone density of the pelvis/proximal femur between individual patients and model the articulation of the hip joints. In contrast to previous surfacebased methods, the ASSIMbased reconstruction approach not only considers the anatomical shape of the pelvis, but also the bone interior density of both the pelvis and proximal femur. The rationale is to use as much information contained in the Xray as possible, in order to increase the robustness of the 3D reconstruction. The acetabular orientation can then be assessed directly from the reconstructed, patientspecific anatomy model.