Rainfall, evaporation, and permeability of water are the most important factors in determining the demand of water. The Daegu area has only a meteorologi observatory and there is not sufficient data for adapting the advanced method for derivation of the estimated of evaporation in the Daegu area. However, by using available data, the writer devoted his great effort in deriving the most reasonable formula applicable to the Daegu area and it is adaptable for various purposes such as industry and estimation of groundwater etc. The data used in this study was the monthly amount of evaporation of the Daegu area for the past 13 years(1960 to 1970). A year can be divided into two groups by relative degrees of evaporation in this area: the first group (less evaporation) is January, February, March, October, November, and December, and the second (more evaporation) is April, May, June, July, August, and September. The amount of evaporation of the two groups were statistically treated by the theory of probability for derivation of estimated formula of evaporation. The formula derved is believed to fully consider. The characteristic hydrological environment of this area as the following shows: log(x＋3)=0.8963＋0.1125$\xi$..........(4, 5, 6, 7, 8, 9 month) log(x-0.7)=0.2051＋0.3023$\xi$..........(1, 2, 3, 10, 11, 12 month) This study obtained the above formula of probability of the monthly evaporation of this area by using the relation: $F_(x)=\frac{1}{{\surd}{\pi}}\int\limits_{-\infty}^{\xi}e^{-\xi2}d{\xi}\;{\xi}=alog_{\alpha}({\frac{x_0+b'}{x_0+b})\;(-b<x<{\infty})$ $$log(x_0＋b)=0.80961$ $$\frac{1}{a}=\sqrt{\frac{2N}{N-1}}\;Sx=0.1125$$ $$b=\frac{1}{m}\sum\limits_{i-I}^{m}b_s=3.14$$ $$S_x=\sqrt{\frac{1}{N}\sum\limits_{i-I}^{N}\{log(x_i+b)\}^2-\{log(x_i+b)\}^2}=0.0791$$ (4, 5, 6, 7, 8, 9 month) This formula may be advantageously applied to estimation of evaporation in the Daegu area. Notation for general terms has been denoted by following: $W_(x)$: probability of occurance. $$W_(x)=\int_x^{\infty}f(x)dx$$ P : probability $$P=\frac{N!}{t!(N-t)}{F_i^{N-{\pi}}(1-F_i)^l$$ $$F_{\eta}:\; Thomas\;plot\;F_{\eta}=(1-\frac{n}{N+1})$$ $X_l\;X_i$: maximun, minimum value of total number of sample size(other notation for general terms was used as needed)