Objective. The time frame for the original CAGE questionnaire is lifetime and it does not quantify drinking frequency and may be less suitable in a population with very few teetotalers. The purpose of this study was to validate a variant of the CAGE questionnaire and compare it with the outcome of a thorough interview according to DSM-III and ICD-10 criteria and to the outcome of biochemical markers in inpatients in a somatic hospital setting. Material and methods. The questionnaire and biochemical markers were tested on a random sample of 130 patients admitted to a department of orthopedic surgery. The result of a diagnostic interview with a trained staff member from the local alcohol treatment unit was used as the gold standard. Data were analyzed by means of receiver operating characteristic (ROC) curves. Results. In this population 25 % had an alcohol problem and the questionnaire proved to be valid, with a sensitivity and specificity of 0.94 and 0.88, respectively, while the positive predictive value (PVpos) was 0.73 and the negative predictive value (PVneg) was 0.98. Carbohydrate-deficient transferrin (CDT) had a sensitivity and a specificity of 0.47 and 0.96, and PVpos and PVneg of 0.80 and 0.85, respectively. Conclusions. This new diagnostic questionnaire is simple, easy to administer and suitable for screening purposes in populations with a high prevalence of at-risk drinkers.