Assessing shoulder mobility is traditionally performed by a clinician using a goniometer, however this method suffers from inter-rater reliability issues. Wearable inertial sensors, image-based systems and 3D-cameras are proposed to objectively quantify shoulder range of motion (ROM). Standardised data collection platforms are required to ensure the consistency of measurements. This paper describes DigitalROM, a Microsoft Kinect 3D camera based system, and a standardised data collection protocol for shoulder ROM assessment optimised for multi-centre clinical trial deployments. DigitalROM is shown to compare very well against image-based ground truth measures (R2= 0.98 and RMSE≤7°) and shows slightly better performance than inertial sensor based ROM measurements (R2= 0.96 and RMSE≤9°). Additionally, DigitalROM offers the ability to ensure patient adherence to the movement protocol during data collection.