Fluticasone propionate dry powder inhaler (FP-DPI) is widely used for the treatment of asthma. However, local adverse effects such as oropharyngeal candidiasis are often seen and mouth washing after inhaling is recommended. In our previous study, we reported a nonlinear relationship between the amount of drug residue and number of times mouth washing was employed. Thus, we developed a compartment model, in which the inhaled drugs were distributed in both easy and difficult to remove areas. Using this model, we analyzed drug removal efficiency in each area with different mouth washing procedures. Three methods of mouth washing were studied; gargling and rinsing in combination, rinsing alone, and gargling alone, following administration of FP-DPI by sprinkling or inhaling. The amounts of drugs recovered from areas considered to be easy to remove (X(1)) and difficult to remove (X(2)) were determined using a nonlinear least-squares program, while the removal efficiency of each of the 3 methods was also calculated. The ratios of X(1) after sprinkling and inhalation were 63.9% and 21.8%, respectively, while those of X(2) were 6.0% and 12.4%, respectively. The numbers of mouth washings required to remove half doses from easy and difficult to remove areas were 0.2 and 1.4 times, respectively, with a combination of gargling and rinsing following inhalation of FP-DPI, while those were 0.3 and 3.6 times, respectively, with rinsing alone, and 0.4 and 5.8 times, respectively, with gargling only, thus demonstrating significant differences among the mouth washing methods for efficiency in the difficult to remove area. The present results show that the employed methods of mouth washing had a significant influence on the removal of drug residues following inhalation of FP-DPI, with gargling and rinsing in combination considered to be the most effective.