Physiological age-related cognitive decline, practice effect and regression to the mean may interfere with the interpretation of psychometric changes between subsequent neuropsychological evaluations. The standardized regression-based (SRB) change score allows investigators to define clinically relevant cognitive change on an individual basis controlling for these confounding factors. We performed a preliminary study to test its applicability and usefulness in the neuropsychological diagnosis of dementia. We derived a regression equation for the tests of a widely used Italian battery for global cognitive assessment, the Mental Deterioration Battery, in a sample of 20 normal elderly and we tested the potential clinical application of the SRB methodology in two cases of questionable dementia.