The biodegradation of toluene was studied in two lab-scale air biofilters operated in parallel, packed respec- tively with perlite granules (PEG) and polyurethane foam cubes (PUC) and inoculated with the same toluene- degrading fungus. Differences on the material pore size, from micrometres in PEG to millimetres in PUC, were responsible for distinct biomass growth patterns. A compact biofilm was formed around PEG, being the interstitial spaces progressively filled with biomass. Microbial growth concentrated at the core of PUC and the excess of biomass was washed-off, remaining the gas pressure drop compar- atively low. Air dispersion in the bed was characterised by tracer studies and modelled as a series of completely stirred tanks (CSTR). The obtained number of CSTR ( n ) in the PEG packing increased from 33 to 86 along with the applied gas flow (equivalent to empty bed retention times from 48 to 12 s) and with operation time (up to 6 months). In the PUC bed, n varied between 9 and 13, indicating that a stronger and steadier gas dispersion was achieved. Michaelis – Menten half saturation constant ( k m ) estimates ranged 71 – 113 mg m − 3 , depending on the experimental conditions, but such differences were not significant at a 95% confidence interval. The maximum volumetric elim- ination rate ( r m )variedfrom23to50gm − 3 h − 1 . Comparison between volumetric and biomass specific biodegradation activities indicated that toluene mass trans- fer was slower with PEG than with PUC as a consequence of a smaller biofilm surface and to the presence of larger zones of stagnant air.