Accurate determination of permeability is critical for the efficient development of coalbed methane (CBM) reservoirs. Permeability determination is mainly based on the feature straight-line analysis method and the plate fitting method requiring pressure build-up data. Although few analytical methods only require dynamic production data, they usually have limitations in application due to complicated solution processes or insufficient consideration of CBM reservoir characteristics. This study improves the previous method for determining the permeability of undersaturated CBM reservoirs. Firstly, mathematical models are established for calculating permeability of undersaturated CBM reservoirs with constant permeability and variable permeability due to pulverized coal output/blockage, matrix shrinkage, and stress dependence during the production process. Secondly, the proposed models are validated by numerical simulation, analytical, and pressure build-up test methods. Finally, the models are applied to the Muai CBM reservoir in Sichuan Basin, China. High-, middle-, and low-permeability zones of the Muai reservoir are classified. The results show that when bottom-hole pressure and water production data of CBM wells during single-phase water flow period are processed by the proposed method and plotted in a rectangular coordinate system, a strong linear relationship can be obtained for CBM reservoirs with constant permeability and variable permeability. The slope of the fitted equations can be used to determine permeability efficiently and accurately. Compared with previous methods, the new method provides better interpretation results for CBM wells with strongly fluctuating production data since the use of the integration of fluctuating data and cumulative water production reduce the effect of production fluctuation on the results. Due to its conciseness, this method can rapidly calculate reservoir permeability around a large number of production wells and obtain permeability distribution in the reservoir, providing guidance for CBM productivity forecasting and development policy adjustment.