The NRCS-CN (Natural Resources Conservation Service-Curve Number) method has been practically applied for estimating the effective precipitation. However, there are no criteria which reflect the geographic characteristics of Korea having more than 70% of mountainous and rice paddy areas, leading to significant errors in runoff calculation. Thus, it is required to estimate the runoff curve number considered Korea land use classification, however there are practical difficulties to conduct the accurate research and experimentation. In this study, after selecting target areas (urban, agriculture, forest), we performed the runoff analysis to redetermine CN values for the selected basins. To do this, curve numbers for soil type A were estimated using genetic algorithm, and then curve numbers for soil type (B, C, D) were estimated using CN aligner equation. Comparing the initial curve numbers with the estimated curve numbers, it was observed that the slightly differences at Chunwang(0), Choonyang(-1), Janggi(-3). Through the above process, this study proposed new curve numbers to reflect observed rainfall-runoff.