Abstract: The simulation research for the test of the artificial rainfall effect with a statistical simulation method on the b:sis of the three standards of the statistical power,accuracy at记sensitivity shows that: (1)the statistical power, accuracy and sensitivity vary with the experimental scheme,statistcal method,natural fluetuation of the rainfall,correlation of the regional rainfall and normal fitting probability,etc; (2) the rando mcross-regression experiment is a new scheme of the highest statistical power,accuracy and sensitivity; (3) the better the correlation of the regional rainfall (r),the higher the statistical power,accuracy and sensitivity; if r is too small (r1/4 and R1/5 are talen.