Background: Appetite control and weight reduction is important for the treatment of chronic disease such as obesity, hypertension, and diabetes mellitus. Visual analogue scales (VAS) is widely used to assess appetite. We investigated the reproducibility and the validity of the Korean version of VAS for appetite which will be helpful for clinical use. Methods: The subjects received the same test meal and 8 VAS questionnaires between 6 weeks. They started to fill out the questionnaire before lunch, continued after lunch every hour, and ended after dinner. The questionnaire was asked about hunger, satiety, fullness, prospective consumption, sweet, salty, savoury, and fatty. During the test meal, the subjects could eat ad libitum until 'comfortable satisfaction'; and after the test meal we calculated energy intake. We assessed the correlation between test-retest VAS for each appetite and evaluated the validity of VAS for hunger with energy intake as "gold-standard". Results: The VAS curves of each appetite were similar between the test and the retest. The VAS of each appetite on the test day was strongly correlated with that on the retest day. The CRs of 4.5 hour mean VAS (20∼34 mm) was smaller than the CRs of fasting VAS (35∼54 mm). The correlation coefficient of Hunger VAS before dinner and the energy intake was 0.436 on the test day and 0.400 on the retest day. The VAS of the sweet was correlated to the total glucose intake (P＜0.05), and the VAS of salty to the salt intake. Conclusion: The validity of the VAS score for appetite, especially hunger, sweet and salty taste was good. Indeed, the reliability of VAS for appetite was good to use this scale in a clinical setting. (J Korean Acad Fam Med 2008;29:736-745)