Carbohydrate deficient transferrin (CDT) is now accepted as a potentially useful marker for the detection of alcohol misuse. It is not clear whether absolute values or values expressed relative to the total transferrin concentration provide the same diagnostic efficiency. CDT was measured in 35 patients with alcohol related liver disease, 35 subjects abusing alcohol without evidence of liver disease and 35 patients with chronic viral hepatitis using two commercial methods (CDTect and %CDT). To compare the methods, results were normalised by dividing the actual result by the upper limit of the reference range. Subtracting normalised %CDT results from the normalised CDTect results demonstrated a linear relationship between CDTect and total transferrin. This linear relationship could be abolished by calculating the CDTect/total transferrin ratio. The sensitivity of the methods was similar with CDTect (43 and 57%) being slightly superior to %CDT (40 and 46%). Specificity was similar (78%) for both methods. Calculation of the CDTect/total transferrin ratio improved the sensitivity and specificity slightly. The linear relationship between CDTect and total transferrin may produce misleading results in populations with a high prevalence of abnormal total transferrin concentrations and could cause difficulties in method comparisons unless taken into account.