This study first employed microscopic plasma simulations to obtain the concentration distributions of key reactive species such as O3 and·OH. These concentration distributions were subsequently introduced into the macroscopic model as dynamic source terms. This approach enhanced the simulation accuracy and predictive capability. A multiscale numerical simulation model of a fluidized bed dielectric barrier discharge (DBD) plasma reactor was then developed. A hierarchical time-scale micro–macro coupling method was applied. The microsecond-scale discharge process was integrated with the macroscopic transport and reaction processes of pollutants. The concentration evolution of reactive plasma species at the microscale and the instantaneous distributions of atrazine and key species in the macroscopic bed were first analyzed. Subsequently, the time-averaged axial distribution characteristics under different gas flow rates were examined. The impact of gas flow velocity on species distribution and reaction behavior was revealed. Finally, the effects of gas flow rate, particle size, and pH on atrazine degradation efficiency were thoroughly investigated. Compared with a fixed-bed reactor, the fluidized bed DBD system provides more efficient gas–solid contact and a more uniform distribution of reactive species. This results in enhanced remediation efficiency.