Distributed soil moisture model was used to estimate daily soil moisture balance for different crops in the Pangani River basin in Tanzania. The model calculates the soil moisture balance using rainfall and reference crop evapotranspiration at grid resolution of 1 km. The reference crop evapotranspiration was estimated using growth stage dependent crop coefficients making the model able to make crop-type sensitive spatially distributed soil moisture balance estimates. The simulated soil moisture balance at each grid was compared to readily available soil moisture (RAM) for a specific crop and a failure of crop was established if the simulated soil moisture was less than RAM for a critical period over which that specific crop starts wilting. Probability of crop failure for each crop were estimated grid wise and mapped to show areas where rain-fed agriculture for specific crop is feasible in the basin. The results shows that beans have probability of failure less than 20% for most parts of the basin. This crop may be considered as drought resistant and suitable for rain-fed agriculture in the basin. The probability of failure of the Maize crop was found to be less than 20% in Maasai steppe, part of the Usambara and areas around mount Kilimanjaro. Rain-fed agriculture is considered feasible only in these areas while areas around North and south pare mountains would not be feasible for rain-fed maize cultivation. The probability of failure for the rice crop was found to be higher than 80% for most parts of the basin and this crop is considered infeasible for rain-fed agriculture in the basin.