The relationships between diet and the occurence of the major diseases of our civilization are of both epidemiological and practical importance to public health problems. Diseases associated with dietary factors rank among the leading causes of illness and death in Korea. Not only for assessment for clinical evaluation, but also for epidemiological concern, we need for rapid and inexpensive methods to assess dietary intake in relating nutrition to health. The validity of a self-administered semi-quantitative food frequency questionnaire was evaluated for a group of 103 men and women aged 26 to 65. Intakes of 8 nutrients computed from the questionnaire were compared with those derived from 3-months diet record. The questionnaire estimates of mean nutrient intake were within about 10% of the mean diet record measurements for nutrients evaluated but iron and vitamin A. The correlation coefficients comparing unadjusted nutrient intakes measured by two methods ranged from 0.27(iron) to 0.57(fat). The correlation coefficients comparing nutrient intakes measured by two methods adjusted for age and sex ranged from 0.26(iron) to 0.50(calories). After adjusted for age and sex or caloric intake, the intakes in iron did not show correlation measured by two methods. A simple and relatively inexpensive questionnaire could be useful information on dietary intake for macronutrients. The modification and revision would be required for assessments of intakes in micronutrients.