The scientific community has made many efforts to produce gridded datasets with high spatial resolution because they are essential for climate change assessment, impact studies, decision-making, etc. This study fits into this context and describes the methods used to prepare a 5-km gridded product of precipitation and minimum and maximum temperatures by merging observed data from meteorological stations, from 1981 to 2016, of Bangladesh, Nepal, and Pakistan with ERA5 reanalysis. The step-by-step methods for station data quality control and the development of the 5- km gridded data are presented. Additionally, we use the 5- km dataset to show the main climate features of the three countries, which facilitates comparison with other data sources in the literature.