The objective of this study is to design multi-layer perceptron artificial neural network (ANN) architecture in order to predict the fatigue life along with different retardation parameters under constant amplitude loading interspersed with mode-I overload. Fatigue crack growth tests were conducted on two aluminum alloys 7020-T7 and 2024-T3 at various overload ratios using single edge notch tension specimens. The experimental data sets were used to train the proposed ANN model to predict the output for new input data sets (not included in the training sets). The model results were compared with experimental data and also with Wheeler’s model. It was observed that the model slightly over-predicts the fatigue life with maximum error of + 4.0 % under the tested loading conditions