The correlation analyses of the body composition were studied on 14 men who were members of our institute as the first experiment for the purpose of evaluation of human nutritional status. The mean age of these men was 21.9 years old. All measurements were conducted at fasting condition in the forenoon. Methods: The basal metabolism was measured by Douglas-bag method. The skinfolds were measured with the Minnesota caliper model at the pressure of 10g/mm2 of the skin surface, and at five sites of the chest, upper arm, upper ileocrista, subscapula and upper knee. The body density was computed from the volume of displaced water by submersed body at the maximal expiration, the body weight in air and the residual lung volume. The volume of displaced water was determined by reading the side burette which was attached to the tank. The error of this method was estimated about ±100ml. The residual air was determined by the forced breathing method of Lundsgaard and Van Slyke. The nitrogen contents of the oxygen gas, expired gas and alveolar air were analysed with the Showlander's apparatus. The body water measurements were made by the antipyrine space method of Brodie and Soberman et al. The serum cholesterol was determined by the modification of Bloor's method. Results: The mean skinfold thickness of six sites was 5.8mm (table 1). This value was only about a half of that in the reports for American (Brozek, Neeman, Pascale). The body density was in the range of 1.0542 to 1.0903 and showed a mean value of 1.0725. The body water (antipyrine space) was in the range of 47 to 65% of body weight and was 57.6% as a mean value. The body fat estimated from the antipyrine space showed the variable values of 19.0, 19.7 and 21.3% of body weight respectively according to each author's equation, and the fat values calculated from the body density by the equations of Keys and Brozek and Pace and Rathhun were 9.0 and 13.0%, as shown in table 2. Such as this, the coincidence was not recognized between the body fat values calculated from the antipyrine space and from the body density. In the correlations between body density and various skinfold thicknesses, the coefficient value was the highest on the chest and were higher on the upper arm and subscapula as shown in table 3. The correlation between body density and the sum of skinfold thicknesses of six sites showed the fairly high value of γ=-0.71. The correlations of the basal metabolism (cal/hr.) to the body weight (kg.), body surface area (m2) and fat-free body mass (kg.) were γ=0.626, γ=0.872, γ=0.874 respectively as shown in table 3. These coefficient values showed that the correlation of the basal metabolism to the fat-free mass was as high as to the body surface. In the correlations of basal metabolism (cal/hr. kg.) to the body density and to the skinfold thickness (the sum of six sites) fairly high coefficients of γ=0.780 and γ=-0.819 were seen. The correlation between the serum cholesterol and the skinfold thickness was not recognized in these healthy subjects.