Abstract: A method and device for reciprocally assigning a spatial position of a patient and/or a part of a patient's body and an acquired, stored patient data set includes producing at least two video image recordings of the patient and/or part of the body from different recording positions via at least one video image source, wherein a three-dimensional position and/or orientation of the video image source relative to a reference coordinate system is known or can be determined. Corresponding tuples of points are identified in the video image recordings, and three-dimensional spatial positions are ascertained and stored from the identified tuples of points. The stored patient data set are matched to a point set of the assigned spatial positions, and a reciprocal assignment or registration of the spatial position of the patient and/or part of the body and the patient data set is determined when the greatest or best match has been achieved.