Solute chemistry and stable isotope tracers of NO 3 -were used to assess bacterial NO 3 -produc-tion and denitrification in a High Arctic glacial ecosystem during 2009. Changes in the NO 3 -con-centration and the d 18 O–NO 3 in all the proglacial streams revealed that up to 95 % of total NO 3 -was most likely bacterially-derived during low flow con-ditions towards the end of the summer (day of year 250). However, overlapping ranges of d 15 N values for snow NH 4 ? , soil organic matter, cryoconite debris and geological nitrogen in host rocks mean that neither the preferred substrate(s), nor the pathway (i.e. nitrifica-tion or simple mineralisation) can be discerned. The most plausible explanation for the bacterial production of NO 3 -is nitrification in snowmelt-fed flowpaths through avalanche fans that flank the glacier and along subglacial drainage pathways at the glacier bed. Interestingly, there was no evidence for denitrifica-tion in subglacial outflow, which is contrary to earlier research at this site. Instead, increases in the d 15 N–NO 3 of up to 20 % downstream of the glacier margin, suggests that denitrification in the glacier forefield and/or the sediments that flank it was most discernable during 2009. Our observations therefore suggest that poorly understood temporal variations in the mixing ratio of nitrifying and denitrifying flow-paths occur in this glacial ecosystem.