Functional green walls are gaining attention due to their air cleaning abilities, however the air cleaning capacity of these systems may be improved through substrate modification. This experiment investigated the capacity of several green wall media to filter a range of air pollutants. Media, consisting of differently sized coconut husk based substrates, and with different ratios of activated carbon were evaluated through the use of scaled down model 'cassettes'. Tests were conducted assessing each substrate's ability to filter particulate matter, benzene, ethyl acetate and ambient total VOCs. While the particle size of coconut husk did not influence removal efficiency, the addition of activated carbon to coconut husk media improved the removal efficiency for all gaseous pollutants. Activated carbon as a medium component, however, inhibited the removal efficiency of particulate matter. Once the substrate concentration of activated carbon approached similar to 50%, its gas remediation capacity became asymptotic, suggesting that a 50:50 composite medium provided the best VOC removal. In full-scale botanical biofilter modules, activated carbon-based substrates increased benzene removal, yet decreased particulate matter removal despite the addition of plants. The findings suggest that medium design should be target pollutant dependent, while further work is needed to establish plant viability in activated carbon-based media.