A prospective pilot study compared the Neurobehavioral Cognituve Status Examination (NCSE) to the Folstein Mini-Mental State Examination (MMSE) to determine the usefulness of the NCSE as a cognitive screen in a geriatric inpatient population. All patients directly admitted to the geriatric evaluation and treatment unit (GETU) of a university teaching hospital over a two-and-a-half-month period were eligible for the study, in which 42% participated, Within 72 hours of admission, patients were given the MMSE and the NCSE in a nonrandom order by a trained psychologist and a structured interview by a psychiatrist. The ability of the NCSE to detect global cognitive impairment was compared to the MMSE and psychiatrist's assessment. Differences in sensitivity were examined by discordant pair analysis. The psychiatrist's determination of the presence of cognitive impairment was used as the criterion standard. Comparison of the MMSE and NCSE, respectively, revealed the following: sensitivity 83% versus 100%; specificity 78% versus 11%; positive predictive value 83% versus 43%; and negative predictive value 78% versus 100%. Seven patients who were cognitively impaired by the NCSE were not impaired by the MMSE (p < 0.05 by discordant pair analysis). The time of administration for the two tests was significantly shorter for the MMSE (14.75 ± 5.7 minutes) than for the NCSE (38.9 ± 12.9 minutes). The NCSE was found to be more sensitive than the MMSE in detecting cognitive impairment among geriatric inpatients, but its specificity and positive predictive values were lower. Beyond this pilot study, additional work examining the utility of the NCSE in other geriatric settings and for different purpose (e.g., as part of comprehensive assessment) needs to be performed.