The design of a dry powder inhaler device has significant influence on aerosol performance; however, such influence may be different between the drug-only and carrier-based formulations. The present study aims to examine the potential difference on the dispersion between these distinct types of formulations, using Aerolizer(®) as a model inhaler with the original or modified (cross-grid) designs. A coupled CFD-discrete element method analysis was employed to determine the flow characteristics and particle impaction. Micronized salbutamol sulphate as a drug-only formulation and three lactose carrier-based formulations with various drug-to-carrier weight ratios 1:5, 1:10 and 1:100 were used. The in vitro aerosolization performance was assessed by a next-generation impactor operating at 100 L/min. Using the original device, FPFloaded was reduced from 47.5 ± 3.8% for the drug-only formulation to 31.8 ± 0.7%, 32.1 ± 0.7% and 12.9 ± 1.0% for the 1:5, 1:10 and 1:100 formulations, respectively. With the cross-grid design, powder-mouthpiece impaction was increased, which caused not only powder deagglomeration but also significant drug retention (doubling or more) in the mouthpiece, and the net result is a significant decrease in FPFloaded to 36.8 ± 1.2%, 20.9 ± 2.6% and 21.9 ± 1.5% for the drug-only, 1:5 and 1:10 formulations, respectively. In contrast, the FPFloaded of the 1:100 formulation remained the same at 12.1 ± 1.3%, indicating the increased mouthpiece drug retention was compensated by increased drug detachment from carriers caused by increased powder-mouthpiece impaction. In conclusion, this study has elucidated different effects and the mechanism on the aerosolization of varied dry powder inhaler formulations due to the grid design.