[en] Purpose: To develop a method of measuring locations of the center of dose in stereotactic radiotherapy relative to the center of the target, and thereby obtain a test of the accuracy of stereotactic radiotherapy (SRT). Methods and Materials: An insert was mounted in an SRT collimator on a 6 MV linear accelerator to provide a photon beam ∼ 1 mm in diameter at isocenter, and a method of measuring radiation center coordinates of arced SRT beams. To simulate a small intracranial target, two halves of a Barium paste column were embedded in two adjacent slabs of a humanoid phantom. A film was placed between the slabs to image the radiation relative to the target center. A surgical head ring and computerized tomography (CT) localizer were attached to the phantom and CT scans were obtained. The scans were entered in a three-dimensional computerized treatment-planning system and radiation isocenter coordinates determined by iteratively moving the 90% isodose surface center of arced beam dose distributions to coincide with the target center. The phantom was bolted to an SRT floorstand with isocenter coordinates obtained from the treatment plan, and then irradiated in two sets of experiments. The first set applied five 1 mm noncoplanar arced beams with and without offsets of the planned coordinates in the transverse plane. The second set applied one large transverse arc coplanar to the film with and without offsets in the craniocaudal direction. Irradiations with coordinate offsets tested the sensitivity of the method. Films were developed and digitized with a high resolution film scanner to measure the location of the radiation relative to the target center. Results and Conclusion: The radiation center was found from 0.0 to 0.3 mm of the target center, within requirements of our clinical quality assurance program. The measurement and evaluation of coincidence of radiation and target centers are, thus, proposed as elements of radiosurgery facility acceptance and annual quality assurance