Background/Aims: The goal of the present study was to evaluate the diagnostic accuracy of the core diagnostic criteria for frontotemporal dementia (FTD) [Neary D, et al: Neurology 1998;51:1546-1554] within a memory clinic population. Methods: The 5 core diagnostic criteria for FTD were operationalised in an informant-based written questionnaire. For a diagnosis of FTD the total clinical picture was weighted with findings on additional investigations and possible exclusion criteria, with follow-up of at least 1 year. Results: The operationalised core criteria for FTD had a sensitivity of 79% (95% CI = 57-92) and a specificity of 90% (95% CI = 85-94). Conclusion: The core diagnostic criteria for FTD applied in a caregiver questionnaire have good diagnostic accuracy among subjects without advanced dementia attending a memory clinic. This stresses the importance of the informant-based history in the differential diagnosis of dementia. Copyright © 2008 S. Karger AG.