Laboratory-scale experiments were carried out using up-flow 7 L Submerged Aerated Filter reactors packed with wool fibre or commercial plastic pall rings, Kaldnes, (70% by volume) support media for the tertiary treatment of sewage. The performance of the wool bioreactor was more consistent than that with Kaldnes medium, for both TOC removal (93%) and SS removal (90%). Both plastic and wool-packed bioreactors achieved complete nitrification at the load of about 0.4 kgCOD/m(3)/day. The sludge yield of the wool bioreactor was almost half that of the bioreactor with Kaldnes suggesting that wool could retain residual organics and particulates. The wool however was degraded and it was concluded that wool would have to be considered as additional sacrificial adsorption capacity rather than an alternative medium. The performance was linked to the residence time distribution studies and these changes in the wool structure. Biomass growth increased the retention of the tracer in the wool reactor by, it was suggested, exposing a greater surface area. Results from the plastic media on the other hand showed increased mixing possibly by increasing the mobility of the plastic. Aeration increased the mixing in both reactors, and patterns were in all cases predominantly well-mixed.