This investigation compared the use of computed tomography (CT) of the head at a large primary medical-school-affiliated hospital and at a large community hospital. There were two aims: first, to study the intrinsic characteristics of the patients in an attempt to determine the potential for developing accurate discrimination algorithms; and second, to study the patterns of neurodiagnostic tests used at these facilities. The results indicated that separability of patients into normal and abnormal categories at both institutions was extremely small. In addition, there was no significant difference in the numbers of types of ancillary tests used at both institutions. Overall, these data once more confirm the difficulty of altering CT usage patterns in primary or secondary hospitals without significantly affecting the number of abnormal patients identified.