Assessing any change in our global climate is a key research area for the future of our population and our environment. Rainfall is a key parameter in the study of the global climate and large scale variations can lead to droughts and floods, two very important environmental phenomena in terms of their impact on human life [2]. Rainfall estimation is one of the most important and challenging operational tasks carried out by meteorological services all over the world. Since 1986, the neural network technique has drawn considerable attention from research workers, as it can handle the complex and nonlinear problems better than the conventional statistical techniques. Several studies have shown the ability of artificial neural network (ANN) in modeling of rainfall estimation. It has the ability to predict future values of the time series from itself. Neural network can be successfully used to predict the chaotic series. Neural network technique is useful both for stochastic and deterministic forecast processes [10]. There two ways of estimate rainfall, that is estimate precipitation (mm) and estimate the number of rainy day. Rainfall estimation for the Penang Island was made based on the 55 years of precipitation per month and 30 years of number of rainy day per month in Penang Island. In this report, a neural network approach for rainfall estimation system will be presented. The neural network based rainfall estimate offers an alternate approach to the rainfall estimation problem. This system is fully software and is written using Neural Network Toolbox in MATLAB. Among the types of neural network that will be developed are Focused Time-Delay Neural Network (FTDNN) and Distributed Time-Delay Neural Network (DTDNN). This study proves that the DTDNN trained using the LM algorithm achieves the best performance as compared to the FTDNN. The DTDNN produces highest accuracy of 66.67% and highest correlation of 0.9047. In this study, an intelligent system is developed for the rainfall estimation using the DTDNN network.