Abstract Background: Carbohydrate-deficient transferrin (CDT) is a marker of alcohol intake that is used for detecting or monitoring alcohol-use disorders. The introduction of a new direct immunoassay for CDT justifies reevaluation of test performance and reexamination of factors affecting test diagnostic sensitivity and specificity. Methods: Individuals enrolled in twin/family studies of alcohol use and dependence provided blood samples and information on recent alcohol use. Serum CDT concentration was measured in 2 088 people with the N Latex CDT (Dade Behring) method, and CDT percentage (CDT%) was calculated as the proportion of the total transferrin concentration measured with Roche reagents. Results: Diagnostic sensitivity was low, both for comparisons of men who reported an alcohol intake of >28 drinks/week vs those who consumed ≤28 drinks/week (28% sensitivity) and for women who consumed >14 drinks/week vs those who consumed ≤14 drinks/week (18% sensitivity), at cutoff values that yielded a 95% specificity. Body mass index, variables associated with metabolic syndrome, and smoking had notable effects on the probability of an abnormal CDT result with excessive alcohol use. Diagnostic sensitivity was greater in men of normal weight (43%) than in obese men (10%) and greater in male smokers (38%) than in male nonsmokers (21%). In women, diagnostic sensitivities were ≤20%, even for those of normal weight and for smokers. Conclusions: CDT is a poor marker of excessive alcohol intake in both women and men who are overweight or obese. It is also less useful in nonsmokers than in smokers. The diagnostic performance of the direct immunoassay and the effects of obesity and smoking are similar to those reported with previous anion-exchange immunoassay methods.