The aim of this study was to investigate the applicability of replacing the glass throat from a twin-stage impinger (TSI) with a human oral-throat cast. Monodisperse aerosols were used to calibrate the human oral cast-TSI at 60 L min(-1) and cut-off in particle size was compared with that of the TSI described in the British Pharmacopoeia which employs a glass throat. The amount of salbutamol sulphate (and lactose) delivered by the Cyclohaler depositing on various elements of the in-vitro model were determined. The calibration of the model containing a human oral-throat cast at 60 L min(-1) gave a particle size cut-off of 5.2 microm which was less than that of the TSI (6.3 microm). The oral-throat cast trapped more drug than the glass throat model with a formulation that employed the larger carrier (63-90 microm; P<0.05) while it trapped a lesser amount of drug with those filled with the lower size carrier (Lactochem, micronised lactose). The greater amount of lactose in the formulation that employed the larger-sized carrier (63-90 microm) was deposited closer to the inlet of the oral-throat cast than to the inlet of the glass throat model. Replacement of the glass throat in the TSI by the human oral-throat cast, leads to a change in deposition efficiency, with the cast having a higher filter efficiency and hence more aerosol particles being captured before their entry into the TSI. This should be investigated further to determine whether such a model might provide a more realistic assessment of the in-vivo characteristics of an aerosol in comparison with the TSI currently being employed, which utilises the glass throat as the portal of entry.