Dry powder inhalers (DPIs) are well established as a means of delivering inhaled drugs to the lungs. Combination DPIs, containing two or more drugs in a single formulation, may improve patient compliance by simplifying medication regimens. The implications of co-formulation on the aerosolisation behaviour of the component drugs, however, are not fully understood. Using salmeterol xinafoate (SX) and fluticasone propionate (FP) as model drug powders, a systematic approach was undertaken to better understand the consequences of combination formulation on drug dispersion for inhaled delivery. A dry dispersion laser diffraction analysis was developed to characterise the inherent dispersibility of bulk powders as well as de-agglomeration of DPI blends. Eight inhaled drug/excipient powders displayed different dispersibility (represented by the DA50, the dispersing pressure to achieve 50 % de-agglomeration; 0.23-1.45 Bar) and cohesivity (represented by the critical primary pressure (CPP); 1.0-3.5 Bar). Upon co-formulation (in the absence of a carrier) SX deposition in the Next Generation Impactor (NGI) became less efficient as the FP content increased (e.g. fine particle fraction (FPF) 33-18% recovered dose (RD) for SX:FP ratios 1:0-1:8). However, FP dispersion was generally unaffected when blended with SX (FPF 26-29 % RD, SX:FP ratios 0:1-8:1). This was attributed to the greater adhesivity and cohesivity of SX and FP, respectively, and changes in bulk blend structure and dispersibility. Intra-batch heterogeneity in particle properties within a bulk powder were also studied using solid state, calorimetric and inverse gas chromatographic analysis. The NGI was used to isolate aerodynamic size fractions which displayed distinct physicochemical and aerosolisation properties. For example, FP sub-populations had better dispersibility (FPF 32 vs. 19 % RD, respectively) whereas an SX sub-population had higher bulk and surface disorder and poorer dispersibility (FPF 21 vs. 33 % RD, respectively) compared to the bulk powders. Upon co-formulation (in the absence of a carrier) the fractions responded differently in terms of their aerosolisation behaviour compared to the unfractionated powders, indicating heterogeneity in the response to co-formulation within a bulk powder. When formulated with a carrier, however, there were no modifications to the dispersion (i.e. FPF) of unfractionated SX or FP upon co-formulation at different SX:FP ratios, however, a fraction of SX co-formulated with FP in a DPI blend exhibited changes to SX dispersion that were comparable to those in the absence of a carrier. The effect of co-formulation on SX and FP dispersion, therefore, was found to depend on the drug ratio, properties of the powder/fraction employed and the presence of a carrier.