In an attempt to design a method which would be more reliable than dietary survey interviews when estimating salt and protein intakes in a population survey, a urinalysis-related study was done on volunteers. Urinary Na, urea N (UN) and SO4, all indices of dietary salt, protein and sulfoamino acids, respectively, were confirmed as reflecting the nutritional condition. Interaction between salt and protein was not observed at least at the dietary levels used in the present study (for salt, 0.33 and 0.1 g/kg body wt/day; for protein, 1.6 and 0.7 g/kg body wt/day). Excretion of components was delayed several days or more after dietary ingestion; nutritional estimation by urinalysis may not be so much affected by daily variables in the diet intake. Partial urine samples proved to have a highly significant correlation with 24-h urine for urinary Na, K, SO4, UN and their creatinine ratios, thus indicating the availability of partial urine samples as substitutes for 24-h urine specimens. Thus, urinalysis is a more readily facilitated, more scientific and more quantitative method for epidemiological nutritional surveys.