Introduction During total hip arthroplasty (THA), accurately predicting acetabular cup orientation remains a key challenge, in great part because of uncertainty about pelvic orientation. This pilot study aimed to develop and validate a technique to measure pelvic orientation; establish its accuracy in the location of anatomical landmarks and subsequently; investigate if limb movement during a simulated surgical procedure alters pelvic orientation. Methods The developed technique measured 3-D orientation of an isolated Sawbone pelvis, it was then implemented to measure pelvic orientation in lateral decubitus with post-THA patients (n = 20) using a motion capture system. Results Orientation of the isolated Sawbone pelvis was accurately measured, demonstrated by high correlations with angular data from a coordinate measurement machine; R-squared values close to 1 for all pelvic axes. When applied to volunteer subjects, largest movements occurred about the longitudinal pelvic axis; internal and external pelvic rotation. Rotations about the anteroposterior axis, which directly affect inclination angles, showed >75% of participants had movement within ±5° of neutral, 0°. Conclusions The technique accurately measured orientation of the isolated bony pelvis. This was not the case in a simulated theatre environment. Soft tissue landmarks were difficult to palpate repeatedly. These findings have direct clinical relevance, landmark registration in lateral decubitus is a potential source of error, contributing here to large ranges in measured movement. Surgeons must be aware that present techniques using bony landmarks to reference pelvic orientation for cup implantation, both computer-based and mechanical, may not be sufficiently accurate.