Constructed wetlands (CWs) treat municipal wastewater through the retention of nutrients and particles. The retention of nitrogen (N) was studied in the laboratory using columns and meso-scale trenches filled with shellsand and light-weight aggregates (LWA). The objective was to examine whether measuring the natural abundance of delta15N in NO3(-) could be used to estimate the relative contribution of denitrification to the total NO3(-) removal in these treatment systems. In both the columns and the trenches it was seen that denitrification was more efficient in shellsand and LWA collected from on-site treatment systems compared to new LWA. This was due to the high pH value (about 10) of new LWA. The enrichment factors (epsilon) from the column study were in general lower than values found in laboratory tests of isotope discrimination in denitrification, but similar to epsilon values found for denitrification in groundwater systems. No enrichment factors could be found for the trench study due to simultaneous denitrification and nitrification and inhomogeneous N transformation patterns. When NH4+ was partially nitrified in the upper parts of the trench, this diluted the 15N enrichment of NO3(-) due to denitrification. Thus, in systems with high NH4+ concentrations and partial aerobic conditions, the method of natural abundance is not suitable for estimating the relative contribution of denitrification to the total NO3(-) removal.