A computer model has been developed which shows that high-impact polystyrene (HIPS) surface roughness (gloss) depends heavily on rubber phase volume and rubber particle size distribution parameters. The model has been developed in a series of steps. First, several tools have been created for isolation and display of rubber particles near the surface of computer-generated resins. Next, a technique for choosing surface points has been devised, using an algorithm which allows the surface to be distributed by any particle near the surface. In step three, a non-linear fit of the surface points produces an abstract surface in the form of a grid. The variability in the array of grid points is a measure of surface roughness. The measured surface roughness of conventional high-impact polystyrene resins correlates to the variables identified by the model. A high percentage of the surface roughness variability has been explained in a correlation using average rubber particle size and rubber phase volume, showing the linear regression approach to be good for prediction of the surface roughness of conventional HIPS resins.