A biofiltration system was tested to remove low levels of acetone from an indoor space. The biofilters were subjected to a range of air fluxes and concentrations of acetone between 100 and 500 ppbv. Passing low levels of acetone through a canopy of green plants did not improve the quality of the air. However, acetone removal by the biofilters with living moss as a principle substrate, reached a maximum of between 1 and 1.6 μmol s-1 m-2 with a loading rate of approximately 2 μmol s-1 m-2. Generally the removal efficiency decreased with increased loading rates over a range of air fluxes (0.05 to 0.2 m s-1) but appear to increase with loading within the slower fluxes. Neither ZERO nor FIRST order kinetics could adequately describe removal. Instead an empirical model that described the natural logarithm of the unloading rate as a function of the natural logarithm of the loading rate and the natural logarithm of the inverse of the air flux fit the data well. The moss coverage appeared to improve acetone removal beyond what would be expected due to an increase biofilter thickness.