This paper presents a computational-fluid-dynamics (CFD)-based Eulerian-Granular approach for characterizing erosion wear in multiphase-flow systems. In contrast to the conventional CFD-based Eulerian-Lagrangian approach, the Eulerian-Granular approach takes account of multiphase dynamics on the basis of the multifluid concept and the kinetic theory; therefore, a more realistic erosion prediction can be achieved. Both the benchmark study and typical applications have demonstrated the effectiveness of the CFD-based Eulerian-Granular approach from dilute to condensed flow systems. Unlike the conventional CFD-based Eulerian-Lagrangian approach and the spreadsheet-based empirical approach, which tend to provide a risky erosion prediction, the CFD-based Eulerian-granular approach is able to capture detailed flow and phase redistribution effects as well as particle/particle interaction involved in multiphase-flow systems. Because fewer assumptions have been made, a more realistic prediction can be expected. The CFD-based Eulerian-Granular approach described in this paper can serve as a general instrument for erosion analysis in multiphase-flow systems, and thus deserves more attention in the erosion community.