Wurster coating is a fluid bed process used during drug manufacturing to coat pharmaceutical beads or pellets. The beads circulate in the container and are coated inside the draft tube with a liquid spray solution. The solvent evaporates after the droplet deposition and a solid layer forms on the particle surface. If two wetted beads collide, there is a certain risk of twinning, which increases at wet process conditions. Thus, a high fluidization air temperature and a low spray rate reduce the agglomeration tendency. Although an increased air temperature enhances the solvent evaporation, it can cause melting or disintegration of heat sensitive materials (core melting) or lead to coating losses. A small spray rate reduces the amount of liquid in the system, but the longer process times increase the manufacturing costs. The goal of this work was the development of a model that predicts the agglomeration risk in dependency on the process parameters. A CFD-DEM model capable of simulating the fluid and particle motion, spray injection, evaporation and temperature in the process was used for this purpose. A coarse-graining model developed for fluidized bed coating processes was applied to reduce the computational effort of the simulations. We introduce two parameters that describe the degree of wetness in the process: (i) the drying time ratio and (ii) the drying efficiency. Additionally, the maximum particle temperature was tracked to avoid core melting. Those parameters were determined for various process parameters and compared to lab-scale coating experiments. It was found that both wetness parameters correlate with the agglomeration tendency observed in the experiments. A systematic variation of the process parameters was performed to evaluate their effect on the process. The agglomeration risk reduces at higher inlet air temperature and flow rate and at lower spray rates. The simulation results were used to identify the optimal set of process parameters, minimizing the risk of agglomeration and core melting. The outcome of this work highlights the benefits of detailed numerical simulations to process design and optimization.