Accurate estimation for groundwater recharge is required for more sustainable management to water resources. On-farm level, deep percolation is one of the water loss sources that researchers and agronomists work hard to minimize. However, on a large scale or district level, it is considered as a valuable source for recharging the aquifer. Indirect estimation to deep percolation would be an economic effective mean, especially for large scale studies. The presented research study aims at checking the capability of Land Information System (LIS) model in estimating deep percolation by soil-water balance equation. The model was validated using precipitation data extracted from meteorological stations randomly selected and distributed along the Nile basin. Multi-thematic map layers including all soil water balance model parameters i.e. deep percolation and change in soil moisture content were developed to the year 2013 at a scale of 10 km2. The relationship between measured rainfall values and estimated ones by LIS has a coefficient of determination (R2) of 0.88. The obtained results revealed a good capacity of LIS model in estimating deep percolation on a large scale from its direct simulation to the soil-water balance parameters. It was found that deep percolation rates were generally higher in the Nile River's downstream than in upstream especially in Egypt, Sudan, and parts of Uganda. This is in contrary to the behavior of precipitation rates and both surface and sub-surface runoff. They were higher in upstream countries than in downstream ones. The highest deep percolation rate (19.63 mm day-1) was observed in Ethiopia and parts of Sudan during May 2013. It was found that low elevated lands with loose texture tend to have higher rates of deep percolation than elevated rocky lands. The current research results could be of great benefit for sustainable water management in Egypt. However, further study should be conducted in different agro-ecological zones in Egypt for more precise model calibration and validation. This may require access to a large amount of field data and long-term meteorological data to run the model.