Abstract: In this stduy, fault analysis of Quantum-dot Cellular Automata (QCA) circuits is carried out employing a proposed hybrid version of ELMAN neural network approach. The QCA are basically designed structures with locally interconnected cellular automata like arrays. The fault analysis is carried out in this study for QCA circuits because of its wide applicability in signal processing applications and related computational applications. The QCA operates in a manner to process information employing a set of dots in a charged configuration module. Considering the QCA design aspect, when the placement of QCA cells at gate level gets altered, this results in reducing the effect of output polarization of the entire configuration. Hence, this study focuses on introducing a hybrid version of ELMAN neural network for performing reliability analysis on the considered QCA circuit. ELMAN neural network is a feed forward recurrent neural network model operating on gradient descent learning rule and its weights are updated using the evolutionary genetic algorithm approach. For testing the given QCA layout for its reliability, the various faults that may occur during the fabrication process are well-noted and analyzed. The proposed hybrid version of ELMAN neural network along with Genetic Algorithm (GA) are applied to the numerous logic gates in QCA module. The proposed model is validated for its effectiveness with the simulation results computed using the QCA designer. Simulation results that the proposed model performs reliability analysis in a better manner in comparison with that of the other methods considered for comparison form the literature.