This paper presents a method that is capable of predicting pore structure of coal-based microfiltration carbon membranes. The tubular carbon membranes are prepared from coal samples with different particle size and distribution by using the extrusion method. According to the relationship between properties of coal samples (particle size and distribution) and pore structure parameters of coal-based carbon membranes, the experiential equations of average pore size and porosity are concluded, and the change law of pore size distribution curve is also summarized. The results show pore structure properties of coal-based carbon membranes can be exactly predicted for different separation requirements by adjusting coal particle size and distribution, which allows us to minimize cost and time consuming experiments in the development of coal-based microfiltration membranes. This paper presents a method that is capable of predicting pore structure of coal-based microfiltration carbon membranes. The tubular carbon membranes are prepared from coal samples with different particle size and distribution by using the extrusion method. According to the relationship between properties of coal samples (particle size and distribution) and pore structure parameters of coal-based carbon membranes, the experiential equations of average pore size and porosity are concluded, and the change law of pore size distribution curve is also summarized. The results show pore structure properties of coal -based carbon membranes can be exactly predicted for different separation requirements by adjusting coal particle size and distribution, which allows us to minimize cost and time consuming experiments in the development of coal-based microfiltration membranes.