This paper employs a model based on Artificial Neural Networks (ANN) to design a CMOS Inverter and Chain of Inverters and determine how accurately the ANN based designs are able to model the complex, non-linear problem of circuit design. ANN is designed to predict the performance parameters of a CMOS Inverter and chain of inverters for a given process technology. A function fitting ANN with Bayesian Backpropagation Regularization as the training algorithm was designed with three hidden layers of sizes 20, 10, 8 respectively. Test performances of 99% were obtained in the various studies performed. These results show that ANNs have a high accuracy and are able to adapt as the complexity of the circuit increases.