Large-scale long-period analyses like NCEP reanalysis have become invaluable for generating robust statistics for model validation and carrying out comprehensive diagnostics. While three strong features of these analyses are their availability for a long period with daily temporal resolution and global coverage, they suffer from coarse spatial resolution and strong model inputs, especially for variables like rainfall. This work evaluates daily NCEP rainfall as a high-resolution analysis through comparison with high-resolution rainfall based on satellite and GTS gauge observations. We use daily composite rainfalls over the Indian summer monsoon domain for the period 2001--2005 from satellite observations available at 10-km resolution and 10-km interpolated NCEP rainfall (I-NCEP). Comparison of observed and I-NCEP rainfall, in terms of parameters like correlation coefficient and phase synchronization, shows that the I-NCEP rainfall provides a good representation of high-resolution observations even at mesoscales and provides a good long-period, high-resolution data set for model validation and evaluation.