A method based on symbolic time series and entropy theory is proposed to analyse the nonlinear behaviours of DC-DC converters. Firstly, the nonlinear continuous system is described by a discrete time series, which is then transferred to a symbol series composed of simple characters; and the series' block entropy is calculated by means of informatics methodology; consequently, a new quantifiable statistical index is obtained. This study takes a one-order voltage feedback DCM and a two-order current feedback CCM Boost converter as examples, and the results illustrate that the coarse-grained statistical method of block entropy, which can quantify the period-doubling and chaos behaviours in DC-DC converters and precisely confirm the appearance of chaos, is a simple and practical analysis method which has not been used in DC-DC converters yet.