To develop an effective, but short, questionnaire to detect heavy drinking, we combined two questions from AUDIT and three from CAGE. We have tested earlier this new Five-Shot questionnaire among male alcoholics who all had a total score of at least 4/7. The purpose of the present study was to find out the effectiveness of the Five-Shot questionnaire among middle-aged men. In the present study, the Five-Shot questionnaire was distributed to all (n = 853) 40-year-old men attending health screening in a Finnish town. The participation percentage in the screening was 70.3%. Self-reported alcohol consumption was used as a gold standard; 557 of these men gave a self-report that they consumed less absolute alcohol than 140 g/week (moderate drinkers), and 70 reported consuming > or = 280 g/week (heavy drinkers). None of the individual Five-Shot questions was superior to any other in detecting heavy drinkers. The cut-off point > or = 3 for Five-Shot gave a 77% sensitivity in detecting heavy drinkers, with a specificity of 83% and an overall accuracy of 83%. The corresponding figures with > or = 2.5 cut-off were 96%, 76%, and 78%. The commonly used cut-off for CAGE, > or = 2, gave a result of 47%, 87%, and 83%, respectively. The area under the receiver operating characteristic curve for the CAGE was clearly worse, compared with that of the Five-Shot. The new Five-Shot questionnaire seems to be efficient in differentiating between moderate and heavy drinkers. It is also easy and fast for physicians to use in different health-care settings. Thus, it can serve as a good aid in early detection of heavy drinking.