The estimation of precipitation using weather satellite is beneficial to measure global rainfall with high temporal and spatial resolution. The understanding of accuracies and limitations of precipitation satellite data is essential to study by verifying the satellite estimation with the observation data from rain gauges. The aim of this study is to verify the radar and rain gauges rainfall data in rainy season (December 2013 to March 2014), which consist three approaches namely point-to-grid, area-weighted and grid-to-grid for daily interval in Ciliwung watershed and point-to-grid for hourly interval in Darmaga and Pondok Betung stations. The verification methods consist of two namely continuous verification statistics and categorical verification statistics. The result of continuous verification statictic shows the highest correlation is point-to-grid and the variability of errors are from 0 to 45.76 according to magnitude of MAE. The radar rainfall are underestimated to observation with ME negatives. The categorical verification statistics shows that accuracies of those three approaches have the average of 0.44 and biases below 1.5. The correlations in hourly interval with point-to-grid approach in Pondok Betung and Darmaga are lower than daily interval, while the accuracies are higher. The application of remote sensing such as radar technology, satellite precipitation estimation feasible to detail rainfall data in a watershed, especially in locations which observation stations are not available.