Air filtering is an effective approach to maintain the indoor air quality while keeping the building energy consumption at an acceptable range. Adsorption filters are one of the most common types of air purifying devices. One concern about these filters, which inspired much research, is to determine their replacement time, since their efficiency decreases over time during the adsorption process. Therefore, a mass transfer model for adsorption filters had to be developed to predict the decay in filter efficiency over time as a function of the bed properties, the air flow rate, and the adsorbent-adsorbate system characteristics. This analytical model is validated systematically with experimental results obtained from a small scale and a large scale experimental setup, for two types of contaminants (MEK and n-hexane), at low, middle and high levels of inlet concentration. The model results are then compared with two previously developed models that solve the equations governing mass transfer; one is analytic and one is numerical. The proposed model shows clear advantages over those ones. Once validated, the model is applied to study the effect of varying four main operating parameters: the convective mass transfer coefficient, the diffusivity within the porous pellets, the air volume flow rate and the pellet size. When studying the effect of varying the air flow rate or the pellet size, the parametric study is carried out in large ranges of Biot number, in order to avoid the influence of convective mass transfer coefficient variations. Indeed it is shown that for large Biot number the controlling parameter is diffusivity within the particles (as anticipated) while, for small Biot number, the convective mass transfer coefficient is the dominant resistance. Finally, the variations of initial efficiency and bed saturation time with respect to changes of these four parameters are discussed.