The present work is aimed at a validation of the carbohydrate-deficient transferrin (CDT) determination in real cases by comparison between the a commercial immunometric method and a method based on capillary electrophoresis. Overall, 650 serum samples from subjects applying to re-obtain their driving license, previously withdrawn for "drunk driving", were investigated. A highly significant correlation (P < 0.001) was found between the results from immunoassay and capillary electrophoresis. However, particularly in the samples with CDT values around the cut-off or moderately elevated, a wide dispersion of the correlation data was found. This finding stresses the need to confirm by alternative techniques all the results from CDT immunoassays. For this purpose, capillary electrophoresis, because of its inherent characteristics of high selectivity, easy operation, high productivity and low operative costs looks well-suited for becoming the method of choice.