The rainstorm frequency curve fitting is essential for the identification of storm intensity formula, the study of rainstorm frequency curve fitting with considering of experience factors was carried out, and put forward to regard the rainstorm intensity frequency curve fitting as an optimization problem, and then to solve it by the weighted damped Gauss-Newton iterative algorithm. Compared to existing methods, the proposed method introduced weight coefficients to improve the fitting precision of commonly used recurrence period in engineering, and to avoid the intersection problem of different frequency curves. The finite difference method is proposed to simplify the calculation of Jacobian matrix, and the damping coefficient was added in Hesse matrix to improve iterative convergence. Thirty-three years of rainfall data of Longyang District of Baoshan city in Yunnan Province were used as an example to illustrate and demonstrate the feasibility and practicability of the proposed algorithm.