Laboratory incubations with varying O2 and NO3 concentrations were performed with a range of filter materials used in constructed wetlands (CWs). The study included material sampled from functioning CWs as well as raw materials subjected to laboratory pre-incubation. 15N-tracer techniques were used to assess the rates of denitrification versus dissimilatory nitrate reduction to ammonium (DNRA), and the relative role of nitrification versus denitrification in producing N2O. The N2O/(N2 + N2O) product ratio was assessed for the different materials. Sand, shell sand, and peat sustained high rates of denitrification. Raw light-weight aggregates (LWA) had a very low rate, while in LWA sampled from a functioning CW, the rate was similar to the one found in the other materials. The N2O/(N2 + N2O) ratio was very low for sand, shell sand and LWA from functioning CWs, but very high for raw LWA. The ratio was intermediate but variable for peat. The N2O produced by nitrification accounted for a significant percentage of the N2O accumulated during the incubation, but was dependent on the initial oxygen concentration. DNRA was significant only for shell sand taken from a functioning CW, suggesting that the establishment of active DNRA is a slower process than the establishment of a denitrifying flora.