High-resolution precipitation datasets are used for numerous applications. However, depending on the procedures for obtaining these products, such as number of observations, quality checking, error-correction procedures, and interpolation techniques, they include many uncertainties. Therefore, the accuracy of these products needs to be evaluated over different regions. In this study, the Iranian National Dataset (INDS), a new 1 × 1 km precipitation dataset based on precipitation data of 1,441 quality-controlled stations for the climatic period from 1961 to 2005, was constructed using the digital elevation model, correlation method, and Kriging interpolation procedure. Iran's annual precipitation values at grids and stations were extracted from Climatic Research Unit (CRU) CL 2.0, CRU TS 3.10.01, and WorldClim datasets, and differences between corresponding values in each of the three datasets and INDS were calculated and analyzed. The coefficient of determination (R2) between the national network stations' data and the CRU CL 2.0, CRU TS 3.10.01, and WorldClim datasets were 0.50, 0.13, and 0.62, respectively. Moreover, R2 values between the grids of each dataset and INDS were 0.51, 0.40, and 0.60, respectively. To determine the global datasets' efficiency for displaying temporal patterns of precipitation, the monthly values gathered from them at 11 stations (as representative of Iran's various precipitation regimes) were compared with the real values at these stations. The results showed that in term of temporal patterns, the concurrences among the three global datasets and the INDS was more acceptable, especially in the case of CRU CL 2.0. In general, it is concluded that the global datasets could be deployed for the primary assessment of the annual precipitation distribution; however, for more precise studies, use of local data is highly recommended.