The urea fluidized bed hydrodynamics were simulated using computational fluid dynamics (CFD) vs. image processing technique. To identify the suitable model, sensitivity analyses on time step, number of nodes, and drag force coefficient were performed, and accordingly, Gidaspow and Syamlal�O�Brien drag models were selected to simulate the fluidized bed for - particles categorized under Geldart-D. The CFD data were validated against the determined particle velocity using an image processing technique. Results indicate that Gidaspow model predictions for particles velocity and granular temperature are in good agreement whereas the Syamlal�O�Brien model resulted in underestimated predictions. Further analysis using Gidaspow model, the particle velocity increased by 12.5, when the air inlet velocity is increased from 2 to 3 m/s and decreased by 87.5 when the particle sizes are changed from 2 to 4 mm. Although the primary loading height and pressure showed no effect, but at high pressures, it is important to consider humidity to avoid the negative impact on the granulation performance. © 2021