Precipitation is one of the important variables in hydrological and water resources studies. Therefore, accurate estimation of precipitation is an important issue. In this study, the performance of three datasets, CHIRPS, ERA5, and PERSIANN-CDR, was evaluated for estimating precipitation in Lorestan Province on daily, monthly, seasonal and annual scales using statistical indices (i.e. nRMSE, RMSE, rMBE, and NSE), as well as precipitation occurrence indices POD, CSI, and FAR. Daily precipitation data from 9 synoptic stations from(2001 to 2022) were collected and compared with the results from the three datasets. Main suppose of this study is that remote sensing and reanalysis precipitation data can be used in locations facing data scarcity or gaps in time series. Results showed that they are reliable at seasonal and annual scales, but at daily and monthly scales, they require bias correction. The results indicated that the PERSIANN-CDR and CHIRPS datasets performed better in various seasons and annually. Based on the NSE and RMSE indices, the best performance was observed from the PERSIANN-CDR dataset in autumn and from ERA5 in spring. Additionally, in winter and on an annual scale, the CHIRPS and PERSIANN-CDR datasets showed better performance. The rMBE index revealed that the CHIRPS and PERSIANN-CDR databsets underestimated precipitation, while ERA5 overestimated it. According to the precipitation occurrence probability indices, ERA5 demonstrated the best performance with an average (POD) of 93% and (CSI) of 82%, and CHIRPS outperformed other datasets with an average (FAR) of 7%. It is recommended to use the seasonal and annual data from these datasets in areas with data scarcity, and to use the daily and monthly data with caution.