||streamflow model output is limited by the quality of the data used to drive, calibrate and validate the model. The most important input to any hydrological model is precipitation, hence inaccuracies in precipitation data are often cited as serious impediments to successful hydrological model. In developing countries, availability of spatially and temporally of precipitation data especially in ungauged basins remain a critical issue. To overcome these problems, this research will review and use some of the available globally gridded high resolution precipitation datasets: (1) Asian Precipitation highly Resolved Observational Data Integration Towards the Evaluation of Water Resources (APHRODITE), (2) Tropical Rainfall Measuring Mission (TRMM), (3) Precipitation Estimation from Remote Sensing Information using Artificial Neural Network (PERSIANN), (4) Global Precipitation Climatology Project (GPCP), and (5) a modified version of Global Historical Climatology Network (GHCN2) to compare with ground station data distributed within Johor River Basin, Malaysia. Statistical analysis such as linear correlation coefficient, mean error, mean absolute error and bias will perform to identify the best precipitation datasets performance. The data downscaling approach will use to improve the spatial variability of the best precipitation dataset. A model will develop to process the best dataset in order to improve the precipitation estimation as input to streamflow model. The new precipitation datasets will input into SWAT, SWMM 5.0 and SOBEK model to simulate streamflow for Johor River. The hydrograph produce by three different streamflow model will compare with discharge data obtain from ground station to identify the most suitable streamflow model for tropical environment. The result of this research can be used for planning purposes such as flood control design and assessment of water resources.