Wireless communications have demand for wireless device of light weight, small size, having low energy consumption and of more practicability. The microstrip antenna is the best choice that fulfills the basic requirements of wireless communication device i.e. antenna should be relatively cheap, light weight, robust, easy to manufacture and with a minimum impact on environment. The microstrip patch antennas are of narrow band width because of which the design parameters of patch antenna must be determined with high precision. So, with in a fabrication process of such kind of antennas, Artificial Neural Networks (ANNs) based models trained with experimental/generated data of same class, previously collected, can preliminary estimate the design parameters, reducing the number of real prototypes to be constructed, saving time and resource. The results obtained using, Artificial Neural Networks (ANN) techniques are close to experimental results and Electromagnetic computer (EMC) simulating tools. Therefore to design Microstrip patch antennas for wireless communication, Artificial Neural Network techniques can be embedded in computer aided design (CAD) system