Background/Aims: Research guidelines on predicting and diagnosing Alzheimer's disease (AD) acknowledge cerebrospinal fluid (CSF) levels as pivotal biomarkers. We studied the usefulness of CSF biomarkers in the diagnostic workup of patients in a geriatric outpatient memory clinic of a community-based hospital, attempted to determine a cutoff age for the use of CSF biomarkers in this group of patients, and compared the total τ/Aβ ratio as an alternative CSF diagnostic rule with the usual rules for interpreting CSF levels. Methods: This was a prospective study of consecutively referred patients. Inclusion criteria were described on the basis of previous study results in the same setting. The CSF tool was applied either to differentiate between AD and no AD or to increase certainty having made the diagnosis of AD. Clinicians were asked to judge whether the CSF results were helpful to them or not. Results: The reasons to use the CSF tool in the diagnostic workup were in 78/106 patients to decide between the diagnosis "AD" and "no AD" and in 28/106 patients to increase the certainty regarding the diagnosis. In 75% of cases the CSF levels were considered diagnostically helpful to the clinicians. Results in the present setting suggest 65 years as the cutoff age to use CSF as a diagnostic tool. The sensitivity and specificity of the total τ/Aβ ratio using the clinical diagnosis as the gold standard were at least as good as the usual categorization rule. Conclusions: Our study results corroborate earlier findings that the CSF tool is of added value to the diagnostic workup in daily clinical practice outside tertiary referral centers. CSF levels can best be used in patients under 66 years of age. Given the limited use of this tool in settings outside research facilities, we recommend that the usefulness of CSF biomarkers is studied in a multicenter study. When in the future CSF levels can be reliably measured in plasma, this may become even more relevant.