The traditional production data analysis (PDA) techniques for gas wells largely relied on the implement of pseudo-functions and related type curve methods. Recently, Ye and Ayala (2012 “A Density Diffusivity Approach for the Unsteady State Analysis of Natural Gas Reservoirs,” J. Nat Gas Sci. Eng., 7, pp. 22–34.) proposed a new rescaled exponential method which can successfully capture the behavior of gas well under boundary-dominated flow (BDF) conditions with the density-based parameters. In this paper, this rescaled exponential method is extended in coalbed methane (CBM) reservoirs by accounting for the two-phase flow behavior, and the variable permeability characteristics in the coal seams that caused by mechanical compression, desorption shrinkage, and the desorption effect. The two-phase rescaled exponential solution for CBM reservoir is derived by modifying the definition of two-phase pseudo-pressure and total compressibility. The proposed rescaled method can evaluate reserve in a convenient and accurate way. Results show that the proposed modification of the original rescaled exponential approach can successfully predict the production of CBM reservoirs compared with the results of commercial numerical simulator (GEM-CMG) and production data of the field cases on the condition of constant bottom hole pressure.