Development of an accurate precipitation dataset is of primary importance for regional hydrological process studies and water resources management. Here, four regional precipitation products are evaluated for the Heihe River Basin (HRB): firstly a spatially and temporally disaggregated Climate Prediction Center Merged Analysis of Precipitation at 0.25° spatial resolution (DCMAP), secondly a fusion product obtained by merging China Meteorological Administration station data and Tropical Rainfall Measuring Mission precipitation data at 0.1° spatial resolution supported by the Institute of Tibetan Plateau Research, Chinese Academy of Sciences (ITP-F), thirdly disaggregated Climate Prediction Center Merged Analysis of Precipitation downscaled by a statistical meteorological model tool at 1-km spatial resolution (DCMAP-MicroMet), and fourthly a weather research and forecasting model simulation with 5 km resolution. The validation metrics include spatial pattern, temporal pattern, error analysis with respect to observation data and precipitation event verification indicators. The results indicate that: 1) precipitation from the DCMAP product may not be suitable for water-cycle studies at the watershed scale because of its coarser spatial resolution; 2) ITP-F, WRF-P and DCMAP-MicroMet precipitation products generally show similar spatial-temporal patterns in HRB but have varying performances between different sub-basins.