Abstract: The accurate evaluation of the joint roughness coefficient (JRC) from 3D laser-scanned point clouds of rock joints is critically dependent on the sampling interval chosen.This study investigates this dependence using rock joint surface samples from Yingshan County,Huanggang City,Hubei Province.A handheld 3D laser scanner was used to acquire point cloud data at 0.1mm resolution.Using a 0.1mm step increment within the range of 0.1~5.0mm,the variation characteristics of JRC with sampling interval for eight rock joints were analyzed by computing the standard deviation of inclination angle (SD_i).Laboratory direct shear tests,uniaxial compression tests,and tilt tests were performed to obtain mechanical parameters of the rock joints.The actual JRC values of the samples were then back-calculated using Barton’s peak shear strength criterion.The results show that the sampling interval significantly affects JRC values,with the trend of change being either positive or negative.At a sample scale of 90mm×90mm,the optimal sampling interval was determined to be 1.0~2.5mm,with an average of approximately 1.9mm.Based on this average interval,five rock joints from the Liujiashan landslide area in Yingshan County were evaluated for roughness.Their cohesion and internal friction angle were estimated using the JRC-JCS model.The analysis indicates that the landslide is stable in its natural state,consistent with field observations.This study provides a new approach to determine a reasonable sampling interval for roughness evaluation of small-scale rock joints,thereby improving the accuracy of roughness assessment using 3D laser scanning.