A mechanical tooth brushing device coupled to an atmosphere pressure ionization ion trap mass spectrometer (API-IT-MS) combination has been developed to study the influence of time and dilution on aroma release from a model dentifrice system. API-IT-MS response to nine commonly used dentifrice flavor components was initially studied. Linear regression models were developed based on an exponential dilution method (EDA) to permit quantification of these compounds. Good linear fits were generated for the majority of compounds (R(2) > 0.92). The threshold detection limits were also calculated, and they greatly depended on the type of aroma compound. A brushing device was then coupled to the API-IT-MS and used to monitor the release profile of three aroma components from a model dentifrice system at flavor concentrations ranging from 0.1 to 20 mg g(-1). Large differences in the aroma release patterns were observed for different compounds (limonene, menthone and cinnamic aldehyde) that depended on their physicochemical characteristics (vapor pressure and log P), and on additional factors such as aroma-matrix interactions. In addition, a linear increase in API-IT-MS response with increased flavor concentration up to 1 mg g(-1) flavor was observed, while at higher concentrations, e.g. between 1 and 20 mg g(-1), a plateau in response was noticed. This suggests that at concentrations above 1 mg g(-1) a transition from a purely dissolved state to an emulsified state occurred. This fact influenced the time-dependent characteristics of the release curve (I(max) and t(max)) for the three assayed flavor compounds.