The aims of the present study are to predict and improve inclusion separation capacity of a six strand tundish by employing flow modifiers (dams and weirs) and to assess the influence of inclusion properties (diameter and density) together with velocity of liquid steel at the inlet gate on the inclusion removal efficiency of a six-strand tundish. Computational solutions of the Reynolds-Averaged Navier-Strokes (RANS) equations together with the energy equation are performed to obtain the steady, three-dimensional velocity and temperature fields using the standard k-ε model of turbulence. These flow fields are then used to predict the inclusion sepapration by numerically solving the inclusion transport equation. To account for the effects of turbulence on particle paths a discrete random walk model is employed. It was observed that with the employment of flow modifiers, the inclusion separation capacity of tundish increases without any large variation in the outlet temperatures. It is shown that inclusion properties and velocity are important parameters in defining the operating conditions of a six-strand tundish.