Introduction <p>Imageless navigation is widely used in total hip arthroplasty (THA), yet evidence for procedures in the lateral decubitus position remains limited because pelvic orientation and registration differ from the supine position. This study evaluated the accuracy of imageless navigation for acetabular component positioning in lateral-position primary THA, using postoperative EOS-based 3D assessment as a postoperative reference method for agreement analysis.</p> Methods <p>The study comprised in-vitro pretests and an in-vivo cohort. In vitro, a pelvic model was systematically rotated along all axes to assess effects on navigated cup inclination and anteversion. In vivo, 70 patients undergoing primary THA in lateral decubitus were included. Intraoperative imageless navigation values were compared with postoperative EOS-3D-measurements.</p> Results <p>In vitro, z-axis (tilt) variations substantially altered both parameters. In vivo, inclination showed a small but statistically significant inter-method difference (mean 1.4°, p = 0.036, Cohen’s d = 0.26), whereas anteversion demonstrated a larger systematic underestimation by imageless navigation (mean −7.5°, p &lt; 0.001, Cohen’s d = −0.78) with poor inter-method agreement (ICC = 0.168).</p> Conclusion <p>Imageless navigation demonstrated acceptable inclination agreement with postoperative EOS assessment, whereas anteversion showed a larger systematic deviation and poor inter-method agreement; sagittal pelvic tilt and positional frame-of-reference differences appear to be major contributing factors.</p> Clinical trial registration <p>The study was registered in the German Clinical Trials Register with the registration number DRKS00026749.</p>