This study examined the utility of the TELE, a telephone assessment for dementia, in a sample of 269 individuals that was not selected on the basis of previous dementia diagnosis. Thus, the conditions of the study reflect the actual situation in which a screening instrument might be employed. Scores on TELE were compared to dementia diagnoses. Using the best cutoff score, sensitivity was .86 and specificity was .90. Longitudinal follow-up established that false positives primarily included those who subsequently developed dementia. Telephone screening for dementia has both clinical and research applications. One recommendation based on our experience is that longitudinal studies should include a telephone interview component for anyone who drops out of the study, to enable characterizing the cognitive status of dropouts.