Remote sensing (RS) techniques have been widely used in water science due to quantitatively and qualitatively development of satellite images. The advantage is pronounced for ungauged basins where climatological, hydrological and qualitative parameters are not available or unreliable. Sentinel-2 multispectral images are applied in this study to estimate the suspended sediment source and concentration (SSC) in flood events for the Laghman River in Afghanistan. Three field deployments (twice in wet and one in dry seasons) were conducted to measure the SSC for calibration purposes. The linear regression model based on Sentinel-2 band 5 (B5) is proposed as a suitable model for predicting SSC in flood events due to its simplicity that provides sufficient accuracy compared to other band combinations. The R 2 and RMSE values are 0.62 and 95.59, respectively. The appropriateness of visible and near-infrared (VNIR) bands for SSC estimation is consistent with similar results in other case studies. Based on the retrieved RS model, the source of sediment production and the spatial variations of SSC along the two upstream tributaries and the main branch of the river are investigated. The results indicate that SSC in Alingar tributary is twice as high as that of Alishing tributary and the concentration is gradually increasing downstream in the tributaries and in the main river. The trend of increase in SSC in the main river is lower than the two tributaries, which indicate that the erosion process is more active in the upstream tributaries than in the main branch.