This document discusses using artificial neural network techniques to develop rainfall-runoff models for the Jhelum River catchment in India. Two types of artificial neural network models were developed - Back Propagation networks and Radial Basis function networks. The radial basis function networks performed better with a root mean squared error of 0.046 and R-squared value of 0.937. The artificial neural network models were able to more accurately model the relationship between rainfall and runoff in the catchment compared to multiple linear regression models. The best performing model was the radial basis function network which emerged as the most suitable for predicting runoff in the catchment area.