THE PURPOSE OF incorporating stereotactic methodology into neurosurgical operations is to achieve a consistently high degree of accuracy in localizing intracranial targets. Therefore, the limits of resolution for the operation are a function of the accuracy of the particular stereotactic frame system. The total clinically relevant error (application accuracy) comprises errors associated with each procedural step, including imaging, target selection, vector calculations, and the mechanical errors of stereotactic frames. To evaluate these parameters, a systematic error analysis was carried out with four commonly used stereotactic devices: the Brown-Roberts-Wells, the Cosman-Roberts-Wells, the Kelly-Goerss COMPASS (modified Todd-Wells), and the Leksell frames. Over 21,500 independent accuracy test measurements were made with 11,000 computed tomograms. The results suggest a potentially significant degree of error in the application accuracy of all stereotactic instruments, which is accentuated by but not entirely due to imaging-associated errors. Clinically encountered levels of weightbearing by stereotactic frames may have a pronounced effect on their mechanical accuracy. Both the reapplication of aiming arc assemblies and the use of phantom base units introduce independent sources of mechanical inaccuracy into stereotactic procedures. The scope of individual error values and their determining factors must be considered with every clinical use of stereotactic frame systems.