A wide-ranging base of jig data was collected from various types of commercial jigs in 9 coal preparation plants in China, and a complete data analysis was made. It was found that the sharpness ofseparation was very variable, which demonstrates the complex interplay of the many factors which influence jig performance. Tests showed that there exist optimum frequency ranges in each compartment of a jig. Stepwise regression analysis indicated that the sharpness of separation is most sensitive to the intake fraction ofthe air cycle, the frequency, and the feed rate. By means of non-linear curve fitting techniques and statistical analysis, several mathematical models were developed to describe both partition curve and the jig performance, for simulation purposes.