Impressive developments in the biomarker diagnosis of Alzheimer's disease (AD and pre-symptomatic states, using cerebrospinal fluid (CSF), positron emission tomography (PET), magnetic resonance imaging (MRI), and proteomics, have been at the forefront of research in the last decade. With the publication of the criteria of Dubois et al. (2007) and the revised National Institute on Aging and Alzheimer's Association (NIA–AA) criteria for the diagnosis of AD (Jack et al., 2011), this has further ignited the interest and enthusiasm for researchers and clinicians in the field. In some of the major conferences on dementia in recent years, the topic of the biomarker-driven diagnosis of AD has dominated the research agenda. But many questions arise as to how this research will translate into practice. Here, we would like to put forward our arguments against a biomarker-driven diagnosis of AD, and we would caution that very early diagnosis of AD may not result in better care of the subjects.