In this paper, two statistical approaches are suggested for the characterization of average and maximum conducted susceptibility in a non-uniform cable bundle. Response surface methodology (RSM) based on Design of Experiment (DOE) is applied for prediction of average current in victim wire within cable bundle. The termination charges at both ends of victim and culprit wires are considered as input parameters to study the current response. The experiments were planned by 4-factor Box-Behnken design (BBD) method. The effect of loads on current response and adequacy of the developed model have been evaluated successfully by Analysis of variance (ANOVA) test. The phenomenon of the occurrence of extreme crosstalk on victim wire and the quantification of extreme risk are studied by extreme value theory (EVT). We demonstrate that EVT can be successfully applied to predict maximum current corresponding to failure probability. In the present study, Generalized Pareto distribution (GPD) using Peak Over Threshold (POT) method was used to estimate the distribution of maximum current and gave appropriate results.