The accuracy of a range of immobilisation systems for head and neck radiotherapy has been measured. Two types of vacuum-formed shells, and eight thermoplastic systems have been evaluated. For each system, a mask was made for the same volunteer. A non-invasive optical system, using a scanning laser digitiser, was used to measure the displacement of a fiducial marker. The volunteer was asked to move a maximum comfortable distance in each of the superior, inferior, left and right directions, and a scan was taken at each location. Standard deviations, calculated from the maximum range of movements, were in the range of 1.7–2.9mm in the right–left direction, and 0.9–3.0mm in the superior–inferior direction. The smallest movements were measured for the Medtec S-frame, with nine-point fixation. A laser-scanning camera can assess the movement potential for a subject inside an immobilisation device with good accuracy and precision. Because ionising radiation is not used as part of the imaging process, the same subject can be used for several mask systems to assess which is the best one.