Dry scrubbing with lime slurry in a spray dryer [spray dryer absorber (SDA)] has been an important technology for flue gas desulfurization (FGD). Mathematical models based on the heat and mass balances are used to predict SO₂ removal in the SDA as a function of initial size distribution of slurry droplets. Since the existence of moisture in the droplets appreciably enhances the SO₂ removal, its removal efficiency depends on the rate of drying as well as that of SO₂ removal itself both depending on droplet diameter. With the increase in the geometric standard deviation (GSD) of the initial droplet size distribution, the efficiency of SO₂ removal first increases and then decreases, showing a maximum at a certain value of GSD. This trend is altered by the sorbent content of the droplets, expressed as stoichiometric ratio (SR). The decrease in SR makes the maximum move to higher GSD and reduces the variation ill the efficiency with respect to GSD. For SR<0.73, a minimum efficiency also appears, ahead of the maximum. The results are well explained by the specific rates of both drying and SO₂ removal of the droplets.