Accurate dynamics of power converters are necessary to achieve good control performance. In this study, the dynamical model of the DC-DC buck converter is identified by the relay feedback method. The relay is connected in the closed loop to produce a limit cycle output. The important information of the oscillatory output is used for the identification. The relay is approximated using dual-input describing function (DIDF) in the mathematical modelling. DIDF can handle symmetric and asymmetric limit cycle outputs. The converter is modelled as a second-order plus dead-time system. Using the gain and phase angle criteria, analytical expressions are derived to estimate the dynamics. The converter dynamics obtained from the proposed method are compared with that estimated using the state-space averaging method. The model is also identified from the realtime experiment. To check the efficacy of the identified model, a model validation test is performed. © The Institution of Engineering and Technology 2018.