The present dataset was developed using the MODIS Normalized Difference Snow Index with a spatial resolution of 500 m as input for the SNOWMAP algorithm to detect lake ice from daily clear-sky observations. Furthermore, for cloud-cover conditions, lake ice was identified based on the spatial and temporal continuity of lake-ice data. On this basis, the daily lake-ice monitoring data of 2612 lakes of the Tibetan Plateau from 2002 to 2018 were calculated and classified. Moreover, a time-series analysis of lake ice coverage, which included lakes with surface area greater than 1 km2, was carried out to provide a clear list of lakes for which lake ice phenology can be estimated. The data set contains 5834 raster files, one vector file and 2612 Excel files (including 1134 time series with and without classification statistics). The raster file is named daily lake ice extent. The vector file contains such information as the