Alzheimer's disease (AD) is increasingly becoming one of the greatest medical challenges. Due to the social and financial burden of AD, detection of AD in its early stages is a topic of major research interest. Thus, emergence of well-validated screening methods for fast detection of AD in the early stages would be of great importance. It is now recognized that the homeostasis and serum bioavailability of multivalent cations (e.g. zinc, copper, and iron) are disturbed in AD. Using a standard chemometric approach (hierarchical clustering analysis), we find that the serum concentrations of an array of such multivalent cations can be a fingerprint for identification of AD patients. This may pave the way for a reliable, efficient, and inexpensive method for early detection and treatment of AD.