The flow in a hydrocyclone classifier is highly turbulent complex nature of multi-phase structure, which has led designers to rely on empirical mathematical models for predicting the performance. A number of classifying cyclone models have been developed and successfully used in mineral comminution circuit simulators over the past four decades (Plitt, 1976 and Nageswararao, 1978). However, most of the mathematical models currently available to describe hydrocyclone classifier performance are based on single average mineral density component behavior. In industrial comminution circuits, the feed to the cyclone composed of a mixture of particles of varying degrees of liberation and sizes. The behavior of multi-component particles in a hydrocyclone is poorly understood and unaccounted for in most of the mathematical model equations available in the literature. There is evidence from a number of experimental data sets indicating the significant influence of the multi-component particles on measured cyclone performance. A concept for modeling the performance of hydrocyclones based on multi-component behavior is proposed in this paper. A set of semi-empirical model equations for cut-size, sharpness of separation are presented and the corresponding model parameters evaluated from the measured data are given.