Abstract Satellite remote sensing is essential for monitoring surface water dynamics on Earth. Space-borne altimeter observations have become an important data source to supplement in-situ measurements of water levels. The applications of satellite radar altimetry in monitoring reservoirs at global or regional scales have been well demonstrated in many previous studies. However, studies on medium- and small-sized reservoirs are limited due to the coarser footprints and relatively low vertical accuracy. In anticipation of new satellite laser altimetry missions, we aim to demonstrate the coverage performance of ICESat-2 for global reservoirs and to further explore its application potential in monitoring the long-term changes in water level and storage of reservoirs by integrating Landsat 8 and Sentinel-2 imagery datasets. In the first 18 months of the ICESat-2 mission, we find that ICESat-2 observations can cover 6,231 out of 7,250 reservoirs worldwide inventoried in the GRanD database with a size ranging from 0.1 km2 to 67166.2 km2, which accounts for nearly 86% in count and about 99% in area and capacity for the whole of GRanD-inventoried reservoir. We then select 40 reservoirs of different sizes and shapes located in different continents to establish the hypsometric curves. Most of these reservoirs show robust fitting in the hypsometric curves, with the R2 values ranging from 0.60 to 0.99 and the RMSE values from 0.37 m to 1.01 m. As a new global satellite altimetry dataset, ICESat-2 shows excellent potential in reconstructing long-term water levels with the hypsometric curve method for various reservoirs. We also find that the ICESat-2 ATL13 product misses a small proportion of reservoirs for various reasons, including their small size, latitudinal narrow shape, and data inconsistency between the HydroLAKES as water mask of the ICESat-2 ATL13 product and the updated GRanD data sets. With the continued observation of ICESat-2 and possibly updated inland water body mask for the ATL13 product in the future, many reservoirs can be routinely monitored with high accuracy. Our findings confirm the powerful capacity of ICESat-2 and are expected to enhance our understanding of reservoir behavior in global hydrological processes and water resource management.