Computerized tomography (CT) continues to be extensively utilized to exclude intracranial pathology in psychiatric practice, but little is known about clinical risk factors, which might predict those patients most likely to benefit from the procedure. We reviewed 150 cases of psychiatric patients who received CT scans to exclude intracranial pathology. We assessed the relationships of patient age, psychiatric diagnosis, and findings from neurologic and cognitive examinations to CT results that influenced patient care, and overall normal and abnormal CT results. Fifty-three percent of the CT scans were abnormal, 11% influenced patient care, and only 2% identified potentially reversible lesions. Cognitive exam results and, to a lesser extent, neurologic exam results, were sensitive predictors of CT findings that influenced patient care. All patients with clinically influential CT results had cognitive deficits and all but one had neurologic deficits. Patients older than 60 years of age and those with organic mental syndromes were most likely to have clinically influential CT findings. Our results suggest that utilizing specific clinical risk factors such as findings from clinical examinations, patient age, and psychiatric diagnosis, to guide the ordering of CT scans, can greatly increase the yield of the procedure for psychiatric patients, without excess medical morbidity.