Compared with the traditional Eulerian-Eulerian two-fluid modeling (TFM) approach, dense discrete phase model (DDPM) is potentially promising for industrial-scale applications of fluidized bed reactors, due to its capability of tracking the trajectories of particle parcels and resolving the particle size distributions (PSD). However, DDPM is yet at its early stage of mature implementation. Hence, verification and validation of the DDPM approach is still required for future applications of different gas-solid fluidization regimes. In the present work, verification and validation of the DDPM is systematically investigated for the first time, with regards to bubbling, turbulent and circulating fluidized beds. The sensitivity of the drag force, a key modeling parameter in gas-solid fluidized bed reactors is investigated with two different drag models i.e., Gidaspow and EMMS/bubbling. Numerical results of bubbling, turbulent and circulating fluidization regimes are presented in terms of discrete particle distributions, axial and radial solid concentrations and radial solid velocity profiles distributions. The results indicate that the sub-grid correction based on the EMMS approach is essential for the DDPM to correctly account for the effects of dissipative structures in all three fluidization regimes. While, the DDPM with the conventional Gidaspow drag failed to do so. Further, validations of the numerical results with the experimental data also prove that the DDPM approach with EMMS/bubbling drag model can resolve the time-averaged axial and radial solid concentrations and radial solid velocity profiles better than the conventional Gidaspow drag model. (C) 2020 Published by Elsevier B.V.