The diagnosis of alcohol abuse based on objective data is a necessary requirement in both clinical and forensic environments. Among the different biomarkers of chronic alcohol abuse, carbohydrate-deficient transferrin (CDT) is world wide recognized as the most reliable indicator. However, several problems about the real meaning of CDT and the reliability of its use for the diagnosis of alcohol abuses are still open, as reported by numerous research articles and reviews. The present article presents a critical review of the literature on CDT appeared in the period from 2001 to 2005 (included). The article is organized in the following sections: (1) introduction, (2) definition and structure of human serum CDT, (3) pathomechanisms of the ethanol-induced CDT increase, (4) preanalysis, (5) analysis, (6) data interpretation, (7) review papers, (8) conclusions. As many as 127 papers appeared in the international literature and retrieved by the search engines PubMed and Scopus are quoted.