1) Children's data (10-18 years old) of back strength, height, grip strength and running long jump from data book of Japan Education Ministry (1964-1981) were analyzed. Back strength was mainly studied and was compared with other data (grip strength etc.) . The groups with high average values for back strength at 10-14 years old did not necessarily show high values for back strength at 17 years old, or vice versa. Values for back strength at 10-14 years old were not significantly correlated with those at 17 years old.
2) Using the data from data book mentioned above, tension of deep back muscle (FMUS) was calculated based on kinetic models (FMUS. I, II, III) . Calculated FMUS values were 3-5 times larger than measured back strength values at each age.
3) Relative change of back strength and FMUS in 1964-1981 were compared. All values for back strength were lower than those for FMUS except in 1967. During the late 1960th and the 1970th, average back strength values gradully decreased, but the decrease of FMUS was less obvious than that of back strength.
4) Based on the data of 422 children (7-12 years old, male and femele), values of diaphragm area were approximated with some assumptions, and then FMUS values were calculated. FMUS values calculated by approximate diaphragm area were significantly larger than those calculated by constant diaphragm area (465 cm2) .
The figure for getting FMUS values easily was offered to avoid troublesome calculation. This consists of two graphs, and one can read FMUS values with reasonable precision. Parameters needed for getting FMUS values are sexuality, height, body weight and back strength.