[Objective] In order to provide data support for sampling method and sampling density selection in regional dynamic monitoring of soil erosion, the influence of sampling densities on estimation precision of soil erosion factors was studied.[Methods] The paper took Mengyin County, in the Yimeng and Tai Mountains national key control areas, as the research object to calculate the precision loss of soil erosion factors estimation under 1% and 4% sampling densities. The precision loss of soil erosion factors estimation under 1% and 4% sampling densities, compared to the soil erosion factors of the county was evaluated by human-computer interaction interpretation, field investigation and statistical analysis, based on the SPOT5 remote sensing images in 2013 and 1:10 000 topographic maps.[Results] ① Under the 1% sampling density, the relative errors of S, E and K factors were 33.48%, 23.46% and 20.64% respectively, which are mainly influenced by slope, land use and soil types, while the relative errors of L, B and T factors were less than 11%. The average relative error of soil erosion factors was 14.44% in the field investigation units. ② Under the 4% sampling density, the relative errors of E, K and B factors were 15.07%, 13.94% and 10.69% respectively, which were mainly affected by land use and soil types, while the relative errors of L, S and T factors were less than 10%. The average relative error of soil erosion factors in the field survey units was 7.89%. ③ The results calculated by the element interpolation extrapolation method under the two sampling densities were higher than that of the grid calculation method. The area of soil erosion under the 1% sampling density was 19.73% higher than that of the grid calculation method, and under the 4% sampling density, it was 11.77% higher.[Conclusion] The 1% and 4% density factors of Mengyin County had different degree of precision loss, which had a certain influence on the results of soil erosion estimation. In the process of dynamic monitoring of regional soil erosion, the appropriate sampling density can be selected according to the demand.