The ability to estimate grinding process responses, like the surface roughness, specific energy, cutting forces and temperature, is essential to understand and optimize abrasive grinding. Several models to predict the process behavior with various input parameters have been published. In all of them, the number of active grits per unit area is an important parameter; however, this parameter has been given limited attention. Due to the difficulties associated with the accurate prediction and measurement of grits per unit area, very few grinding models have considered this factor, and most treat it as a constant. The active grits count depends on the depth of cut, the type of wheel, and the nature of the workpiece material. In contrast, the cutting edges count depends mainly on wheel-dressing conditions. A method has been developed to predict the active grits count by assuming a stochastic distribution of abrasive grits and their physical interaction with the workpiece based on a presumed abrasive wear mechanism. In addition, the count of active cutting edges has been included by using dressing kinematics. The variation in the number of active grits count with the depth of cut, grit size, dressing conditions and work material nature was used to validate the efficacy of this newly-proposed, predictive method. © 2017 Elsevier B.V.