In this paper, second of the two-part series, pressure-permeability trend established in the laboratory was compared with modeled results using two analytical models, Shi and Durucan and Palmer et al., developed to predict changes in permeability of coalbed methane reservoirs with continued production. The experimental results did not match the modeled results very well for either of the two models. Also, there appeared to be several uncertainties, primarily the constancy of cleat compressibility and estimation of certain input parameters. A modification of the Shi and Durucan model was then proposed to increase the impact of the shrinkage effect on permeability. With this modification, the measured and modeled results matched perfectly, with the modeled permeability of coal increasing continuously, just as observed in the laboratory and supported by field data, and the rate of increase accelerating at low pressures. The phenomenon was explained by analyzing the permeability as a function of reservoir pressure. (C) 2011 Elsevier Ltd. All rights reserved.