Landslides are typically triggered by earthquakes or rainfall, and occasionally slope changes occur due to the continuous effect of rainfall or snowmelt. Significant amounts of sediment from earthquake‐induced landslides usually flows downstream for a long period of time. However, the dynamics of recently eroded sediments from the landslide bare slope are poorly understood. This study initially verifies the accuracy of the terrain model derived from Unmanned Aerial Vehicle‐Structure from Motion (UAV‐SfM) survey on the landslide bare slope by comparing it with the topographic model derived from a Terrestrial Laser Scanning (TLS) survey. The UAV‐SfM survey derived a terrain model with a high resolution of approximately 10cm. Further, due to the effects of 2018 Hokkaido Eastern Iburi Earthquake, sediment dynamics on the landslide bare slope was analyzed for two consecutive years at frequency of every 3months using UAV‐SfM survey. Immediately after the earthquake, unstable materials on slope collapsed. Subsequently, the sediment dynamics exhibited seasonal fluctuations. Snow glide and rock slaking caused surface erosion of sediments during the winter. In the snowmelt season, gully erosion occurred due to snowmelt flooding. The eroded sediments formed talus cones at the foot of the slopes. In the summer (the first year after the earthquake), gully erosion occurred not only on the bare landslide slope but also on the talus cones due to the surface flow of rainfall. During the winter and snowmelt season (the second time after the earthquake), surface and gully erosion occurred again on the lower side slope, whereas gullies on the talus cones were filled with debris supplied from the bare slope. In consequence, however, the gully erosion of the talus cones occurred by the rainfall. This implies that the eroded sediments on the landslide bare slope are discharged downstream during the snowmelt period with a time gap. Thus, this study provides an in‐depth understanding of sediment dynamics on landslides caused due to an earthquake, using multi‐temporal UAV‐SfM survey.