There is growing consensus that it is possible to diagnose probable Alzheimers disease (AD) in patients prior to the development of mild dementia, which was highlighted by a recent international consensus proposal for new research diagnostic criteria (Dubois B, et al., Lancet Neurology, 2007). Data are accruing from longitudinal studies to support the concept that a specific mild amnesic syndrome, detected through clinical assessment of symptoms and signs, is the most common early manifestation of prodromal AD. Whether or not patients meet typical clinical trial criteria for mild cognitive impairment (MCI), the presence of this syndrome is a very sensitive indicator of AD pathology and elevated likelihood of imminent mild dementia. Diagnostic specificity can be improved in this population through the detection of biomarkers typically associated with AD, including neuroimaging measures of abnormal brain structure and function, as well as cerebrospinal fluid markers. It now appears reasonable, using currently available methods, to diagnose probable AD in individuals who manifest this multidimensional clinicobiologic pattern of abnormalities. Since the likelihood of mild dementia is greatly elevated in these individuals, it is critical to begin to translate this knowledge into standard practice in clinical research with the aim of targeting this population for clinical trials of potential disease-modifying therapies aiming to prevent mild dementia. In this manuscript, evidence for these points is reviewed and a proposal is outlined for translation of this knowledge into a research program.