Mild cognitive impairment (MCI) is a nosological entity proposed by Petersen in 1999 with the objective of identifying an early stage of dementia. The new diagnostic criteria for dementia, both those promoted by the International Working Group (IWG) and those defined by the National Institute of Aging (NIA), subsequently introduced a new model that starts with a preclinical phase, then proceeds with a prodromal phase, and ends with a phase of dementia. The condition known as subjective cognitive disorder (SCD) is included between the preclinic and the prodromal phases. Most clinicians improperly consider MCI and SCD as diseases, and not as risk factors for dementia. This ambiguous scenario requires the application of a public-health standard. A diagnosis of either SCD or MCI comes with several uncertainties, raising issues pertaining to both the research setting and clinical practice. A large proportion of subjects with either SCD or MCI will never progress to dementia, and part of them may even revert to a normal cognitive profile. Thus, communicating of these diagnoses to a subject has ethical implications that cannot be underestimated. The frequency of these diagnoses in general population is starting to show the characteristics of both the phenomena of over-diagnosis, and consequently over-treatment. Moreover, the new criteria require the use of biomarkers, that are not yet validated for the use in clinical practice. No population studies are currently available performed based on the new diagnostic criteria for dementia. This means that future estimates will probably be more than twice the current ones, and will include also subjects that will not progress to dementia. This undefined framework, thus, urges the implementation of public-health programs aimed at both the primary and secondary prevention of dementias. Moreover, clinical trials on drugs in MCI currently use endpoints based on non-validated biomarkers, thus raising the issue of the external validity. A public guideline would thus be crucial to support clinicians in identifying all the possible causes that can determine a cognitive decline people that are mainly elder, with multiple comorbidities, and taking multiple medications. Moreover, the widespread use of neuro-psychological assessment tools in current clinical practice requires an accurate validation of these instruments.