Purpose Precipitation data are an important input variable for soil loss (SL) and sediment export (SE) modeling, although they are rarely available at fine spatial and temporal resolutions. As such, in this study, we aimed to evaluate the performance of three high-resolution satellite precipitation products (SPPs) (CHIRPS v02, GPM IMERG v06, and TRMM 3B43 v07) for SL and SE prediction in the Tamor River basin (TRB), Nepal. Materials and methods We conducted our evaluation by comparing SPPs directly with in situ gauge precipitation data and indirectly through SL and SE simulations using the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model. The Nash-Sutcliffe efficiency (NSE), percent bias (PBIAS), root mean square error (RMSE), and correlation coefficient (R) were applied to assess the performance of the SPPs in different spatiotemporal dimensions. Results and discussion Results demonstrated that the CHIRPS v02 product performed the best across all spatiotemporal scales and for all evaluation techniques. CHIRPS v02 also performed the best in the middle- and high-elevation categories in the study area, followed by TRMM 3B43 v07, whereas TRMM 3B43 v07 outperformed the other two products in the lower elevation ranges (< 700 m.a.s.l.). In the temporal scale analysis, all three SPPs performed better at the monthly time scale (R = 0.82 to 0.85), but their performance declined to an unsatisfactory level (R = 0.10 to 0.54) as the temporal resolution increased (monthly > seasonal > annual). In the SL and SE model simulation evaluations, all three SPPs performed better (R > 0.86, NSE > 0.83), regardless of elevation difference. Conclusions Taken together, the CHIRPS v02 product has wider application potential over GPM and TRMM products for SL and SE modeling in the TRB and can be replicated for other basins with similar geographical conditions.