Nitrogen is a key nutrient stimulating algal blooms in Seqwater reservoirs throughout southeast Queensland (Muhid and Burford 2012, Muhid et al. 2013, Burford et al. 2014). Previous collaborative work has shown that management strategies, such as re-planting riverbanks, are likely to have limited capacity to reduce nitrogen (N) loads in Southeast Queensland rivers (e.g. Olley et al. 2015), therefore other mechanisms are needed to remove or retain N. One mechanism is to use the internal processes within the river channel to remove nitrogen, including the use of denitrification. This process involves bacterial conversion of nitrate (a key N source from catchments) being converted to N gas which is exchanged with the atmosphere. Previous studies in constructed wetlands have shown that denitrification can be efficient at nitrogen removal if an environment is created which is suitable for the denitrifiers (Lee et al. 2009). Therefore, this study examined the role of in-channel denitrification for N removal in the Upper Brisbane River, and assessed whether there is scope to enhance these processes during low flow conditions. A series of experiments were conducted between December 2015 and August 2016. Denitrification was measured in sediments with and without aquatic plants (macrophytes) at six sites covering a range of sediment textures (71-98% sand/gravel, grain size >20 μm, and 0.1-2.0% clay, <2 μm). Denitrification was quantified using intact sediment cores via the isotope pairing technique which provides a more accurate estimate of in situ denitrification than previous methods used in southeast Queensland rivers (Udy et al. 2006). The key findings of this study were: 1. Denitrification rates were typically higher in sediments of macrophyte beds (mean ± standard error, 0.49 ± 0.2 mmol N m-2 h-1) compared to bare sediments (0.19 ± 0.09 mmol N m-2 h-1) (higher at 4 out 6 sites sampled). Overall, denitrification was significantly more efficient at removing water column dissolved inorganic nitrogen in sediments vegetated with macrophytes (13.7 ± 5.78%) compared with bare sediments (3.1 ± 1.74%). 2. Overall, denitrification rates ranged from below detection limits to 2.7 mmol N m-2 h-1 (0-70% denitrification efficiency). Denitrification rates at the upper end of this range were comparable with other systems, while rates measured at half the study sites were low compared to other systems, suggesting there is scope to improve N removal in the Upper Brisbane River. 3. Based on denitrification efficiencies estimated in this study and gauging station flow data, up to 0.025 kg N m-2 d-1 (9.1 kg N m-2 yr-1) could potentially be removed from the water column at the Gregors Crossing site under 100% cover by macrophytes, compared to only 0.008 kg N m-2 d-1 if no macrophytes were present. 4. The build-up of organic carbon in sediments associated with macrophytes may have contributed to the enhanced denitrification rates observed in this study. Total organic carbon concentrations were higher in macrophyte sediments compared with bare sediments at four out of six sites. High denitrification rates were also associated with higher sediment total nitrogen and a higher proportion of clay (grain size <2 μm). However, further investigation is required to clarify these relationships 5. Management of the Upper Brisbane River to increase macrophyte cover is likely to enhance N removal via denitrification under low flow conditions. Potential management options could include fencing off river banks to prevent stock access, therefore reducing macrophyte loss through cattle trampling and reducing losses of suspended sediments that smother plants. Other methods which reduce flow velocities during high-flow events would also prevent macrophyte loss through scouring. Re-establishing lost macrophytes post flooding may also be achieved through active planting. Planting a mixture of macrophytes species would increase resilience to variable flow conditions and invasion by exotic species.