Having an accurate understanding of the scale effect of surface morphology characteristics is crucial to examining the mechanical behavior of rock structural plane. At present, the quantification and sampling methods of surface morphology show diversity, which is the potential reason for the inconsistent research conclusions on scale effect. Firstly, based on mathematical statistics and correlation analysis, the most representative parameter is proposed from hundreds of morphological parameters. Then, the previous scale effect sampling methods are analyzed. In order to ensure that the selected samples are representative, a novel sampling method, considering all morphological information, is proposed. By means of the novel quantification and sampling methods, the size effect characteristics are systematically analyzed. Under the conditions of different rock types, shear directions and sampling locations, etc., discontinuity roughness does not change significantly with sampling scale. As sampling scale increases, the distribution range of representative samples is gradually concentrated, the total amount decreases, and the proportion increases. However, the distribution of representative samples on the initial structural plane does not show obvious regularity. These findings would provide theoretical support for the deformation control and stability analysis of rock mass in engineering.