We developed a self-administered diet history questionnaire (DHQ) for use in prevention and control of cardiovascular diseases and cancer, and validated it by comparison with single 24-h urinary excretion of sodium (Na) and potassium (K). The subjects were 154 male and 69 female freshmen university students. Mean intakes (mmol/day) assessed by DHQ and the urinary excretion of Na were 196 and 165 respectively for men and 179 and 136 respectively for women. Those of K were 61.5 and 43.9 respectively for men and 56.8 and 41.6 respectively for women. The ratios of urinary excretion to dietary intake of Na were 0.97 in men and 0.84 in women. Those of K were 0.78 in men and 0.80 in women. The results for both Na and K were reasonable, except for Na in men. When Pearson correlation was examined between dietary and urinary Na and K, no significant correlations for Na in men (r=0.14) or women (r=0.23, p=0.06), or significant correlations for K in men (r=0.34, p<0.001) or women (r=0.40, p<0.001) were observed. The results suggest a reasonable ability to estimate a subject mean for Na in women, K in both sexes, and individual level for K for both sexes. The validity for individual level for Na intake is not conclusive because the duration of urine collection was too short. (Jpn Circ J 1998; 62: 431-435)