This paper presents some results of a study,in which the data associated with typhoons during years of 1974-1996 from Ocean Data Buoy Stations of Japan Meteorological Agency are processed by using statistical methods.As the first part,there are mean values of wind speed,wave height and period in each class of data which is yielded by classifying the data according to typhoons central pressure P0 and the distance from observation point (Buoy) to the typhoon center r and the azimuth angle from the direction of movement of the typhoon center.Then,in order to derive the empirical relationship among the significant wave height,wind speeds and other parameters under typhoons the regression analysis and dimen-sionless analysis are carried out to deal with two sets of the severest data composed from several classes of the classified data respectively.Two proposed equations derived by dimensionless analysis may be used to predict conveniently significant wave height due to typhoons by using the wind speed V10 and the distance r. The equations are tested by previously published field data in South China Sea and found to be in general agreement.