Radiostereometric analysis (RSA) is a stereo X-ray technique used in clinical research studies to evaluate micro-motion and wear of orthopaedic implants within bone. While highly accurate and precise, its adoption has been limited due to technical requirements such as the need for implanted marker beads and radiograph view angles determined by a calibration cage. We propose a new technique that separates the calibration procedure from the patient examination, enabling clinical radiograph views to be used for RSA measurements. The concept of a reference plate was adapted to establish the link between calibration procedure and patient examination procedure for cassette radiography, which may not be necessary for digital radiography. A hip wear phantom was used to validate this technique by comparing the error and repeatability of the novel procedure with that of conventional RSA. Femoral head penetration was measured versus the acetabular cup (head/cup) and marker beads in the acetabular liner (head/liner). Conventional RSA had lower inferior-superior average error (p = 0.03 for head/cup) while the modified RSA had lower anterior-posterior average error (p = 0.01). Average error was greater but not significantly so for the medial-lateral (p = 0.06) and 3D (p = 0.97) measurements. The head/liner method had lower average errors (p < 0.0001) for both procedures, but did not affect repeatability, which was similar between techniques. The novel procedure's average error and repeatability was therefore, similar to conventional RSA. This new technique could be applied to any joint with two clinical radiograph view angles pending further validation in subjects. © 2016 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 34:1521-1528, 2016.