<p>With advancement in remote sensing and climate modeling, several global rainfall products have been developed to supplement in-situ rain gauge observations, especially in data-sparse regions. Evaluating the accuracy and reliability of these products in reproducing local rainfall patterns over the Fiji Islands is key, particular, given the country’s exposure to extreme weather events such as cyclones, drought and floods. Accurate rainfall data are crucial for reliable applications in hydrological modeling, flood and drought monitoring and forecasting, disaster risk reduction, and climate-resilient planning in this tropical island environment. This study evaluates the performance of 15 global rainfall products (6 station-based, 4 satellite-based and 5 reanalysis products) against ground-based observations over the Fiji Islands for the period 1983–2016. Several performance metrics including standardized mean difference (SMD), coefficient of determination (R²), percent bias (PBias), root-mean-square error (RMSE) and the Nash-Sutcliffe efficiency (NSE), are used to evaluate the products. Overall, the evaluation reveals that the best-performing products are station-based rainfall products (REGEN_ALL_V1_2019, GPCC_FDD_v2020, GPCC_FDD_v2018, GPCC_FDD_v2022). These products performed best across lowland stations on the larger islands, whereas reanalysis (ERA5) and satellite-based products (PERSIANN_v1_r1, CHIRP_V1 and CHIRPS_v2.0) showed better agreement with observed data over smaller and isolated islands. However, none of the evaluated products accurately reproduced rainfall for Monasavu station located at high altitude. These findings highlight the preference for station-based products, in providing reliable rainfall inputs for climate services, hydrological modeling, and climate-resilient planning in the Fiji Islands.</p>