<p>To quantitatively characterize the complexity of coal pore structures and establish an intrinsic relationship with gas desorption characteristics, this study utilized low-pressure nitrogen adsorption experiments to conduct a series of experimental investigations. The primary factors influencing the complexity of coal pore structures were identified using the Pearson’s method, and a model was developed using a multiple nonlinear fitting method to represent this complexity. Subsequently, the study explored whether the initial gas desorption rate was related to this complexity. The results indicated that fractal dimension can serve as a bridge for representing pore complexity because it had strong correlations with pore volume, specific surface area, pore-throat ratio, and tortuosity, per their correlation coefficients of 0.466, 0.809, 0.912, and 0.553, respectively. The correlations among porosity, pore-throat ratio, and tortuosity were relatively weak, and so these three parameters were considered as the main independent factors. A multi-parameter model was jointly established to comprehensively characterize the complexity of the pore structure, and the new model overcame the shortcomings of previous single-parameter representations. As the complexity of the pore structure increased, the initial rate of gas desorption decreased. The correlation between the initial rate and complexity for all coal samples was 0.453. After removing outliers, the correlation reached 0.793. This confirmed the accuracy and rationality of the pore structure complexity model in representing the initial gas desorption rate.</p>