Abstract: The research of the harmonic contribution is helpful to improve the power system harmonic management. Previous harmonic contribution research focuses on the harmonic contribution indices and the calculation method for sample harmonic contribution, did not involve the calculation method of the harmonic contribution statistical characteristics, which is the key index of power quality economical evaluation. In the paper, kernel density estimation method was proposed to approximate probability density function firstly. Secondly, the importance sampling Monte Carlo method was adopted to realize the harmonic contribution sampling. Finally, the function relationship between the harmonic contribution statistical characteristics and the harmonic contribution samples was established to realize the statistical characteristics calculation. The simulation and the field test show that the expectation and variance results from the harmonic contribution samples are more accurate and suitable for the power quality economical evaluation, compared with the results from the original harmonic contribution data.