Intraoperative patient registration and tracking are key enabling components of Computer-Assisted Orthopaedic Surgery (CAOS). Optical tracking systems that detect rigidly-attached reference frames with reflective markers are the standard in current CAOS systems, due to their large field of view and high accuracy. Recent improvements in depth sensing technology have led to the widespread availability of RGBD stereocameras. Their adoption in CAOS is appealing, as it would allow to reconstruct a patient’s anatomy with a simple ‘snapshot’ of the scene. However, integrating such cameras into a CAOS system would not be straightforward, as would require changes to the layout traditionally used in current systems. Indeed, while the camera is normally positioned laterally to the patient, a markerless system would need to look directly into the surgical incision, at a much closer distance so as to maximise the density of points in the Region-of-Interest (RoI). While these constraints present some challenges, the light weight and small form factor of many state-ofthe-art RGBD cameras allow to envision solutions that would not be possible with bulkier systems. In this abstract we present a preliminary study for an integrated solution composed of a lightweight RGBD camera connected to the robot-assisted burring tool of a commercial CAOS system. We present preliminary results on the tracking accuracy of the system, and discuss the advantages and disadvantages of such a solution.