Total ammonia nitrogen (TAN) removal efficiencies of a sand filter (SF), polystyrene microbead filter (PF), and Kaldnes bead filter (KF) media were evaluated under ammonia loading rates of 5, 25, and 50 g $m^{-3}day^{-1}$. The volume of each filter media tested was 7 L, and the water flow rate for all filter media was 24 L/min. The specific surface areas of the SF, PF, and KF were 7,836, 3,287, and 500 $m^2/m^3$, respectively. Sand was fluidized and the other two media were trickle filtered. The volumetric TAN removal rate increased with increasing ammonia loading rate for all filter media. Mean volumetric TAN removal rates under the ammonia loading rates of 5, 25, and 50 g $m^{-3}day^{-1}$ in SF (39.3, 168.6, and 322.7 g $m^{-3}day^{-1}$, respectively) were higher than those in PF (35.0, 157.4, and 310.5 g $m^{-3}day^{-1}$, respectively) and KF (32.1, 142.5, and 288.1 g $m^{-3}day^{-1}$, respectively). These results were related to differences in the specific surface areas of the filter media. PF was the most economic media for efficiently removing TAN.