Five different grades of lactose namely, anhydrous lactose, medium lactose, regular lactose, lactose crystals and foremost lactose were fractionated under similar conditions to obtain a size range of 63-90 μm and were characterised using laser diffraction and time-of-flight particle sizing techniques, scanning electron microscopy, optical microscopy image analysis, thermal gravimetric analysis and differential scanning calorimetry. Each of these lactose fractions were then blended separately with micronised salbutamol sulphate in a ratio of 67.5:1 (w/w). The mixing uniformity and percentage recovery of salbutamol sulphate in the powder blends were analysed using a validated HPLC method. The deposition profiles of the drug were determined using a 5-stage liquid impinger after aerosolisation at 60 l min-1 via a Rotahaler. Despite the identical processing conditions, the lactose fractions were shown to differ in particle size, size distribution and concentrations of fine particles. The particles from each fraction also exhibited different surface textures and dissimilar DSC thermograms. However, all the blends of the lactose with salbutamol sulphate were found to have a relatively high uniformity of salbutamol sulphate content, as suggested by a coefficient of variation of less than 3.2%. Anhydrous and medium lactose produced a more efficient delivery of salbutamol sulphate when aerosolised from the Rotahaler in comparison to other grades of lactose. For example, the fine particle fraction (FPF) and fine particle dose (FPD) of drug from formulations containing anhydrous lactose were 13.4±4.2% and 57.3±17.6 μg, respectively, which were approximately two times higher than the respective values of the formulation containing regular lactose. Medium lactose resulted in drug FPF (7.9±2.7%) and FPD (32.4±11.8 μg), which were significantly (ANOVA P<0.05) higher than the same parameters obtained using lactose crystals, foremost lactose and regular lactose. More efficient drug delivery from anhydrous lactose may be partly attributed to the relatively higher concentration of fine lactose in this grade of carrier, although it showed a rougher surface than the other grades of lactose. However, the relatively high FPF of the drug from medium lactose may have been due to the relatively small mean particle size and smooth surface of the particles. Therefore, the source and grade of lactose may have a substantial effect on drug delivery from dry powder inhaler formulations and care should be taken in establishing appropriate quality control parameters when selecting an appropriate grade of carrier.