ABSTRACT: Isotope pairing (IP) and N₂:Ar techniques for measuring sediment denitrification were compared in muddy sediments in a shallow sub-tropical embayment. At a concentration of 100 µM ¹⁵NO₃^– (our experimental conditions), IP greatly underestimated denitrification, compared to N₂:Ar, in low-respiration sediments subject to bioturbation, where denitrification produced via coupled nitrification-denitrification (D_n) dominates overall denitrification. Consistent with this underestimation was the higher N₂ flux measured using N₂:Ar ratios, accounting for more of the stochimetrically predicted N mineralisation. We suggest that this underestimation is primarily due to non-homogenous mixing of the added ¹⁵NO₃^– with the endogenous ¹⁴NO₃^– in the sediment, associated with differentials in the rate of diffusion of ¹⁵NO₃^– across mucus linings in burrow structures versus rates of ¹⁴NO₃^– supply from nitrifiers in the burrow walls. Strong support for this hypothesis was provided by the relationship between the methods discrepancy and polychaete biomass (r² = 0.99; p = 0.001), and the almost complete removal of the methods discrepancy when the upper heterogenous sediment layer that contained the burrow structures was experimentally removed. Due to a significant increase in potential denitrification sites in association with burrow structures, this effect can potentially cause a large underestimation of denitrification (up to 85%) using the IP technique. Our data show no benefit in longer pre-incubation times, suggesting that this effect may represent a fundamental limitation of the IP method in the sediments of oligotrophic (low carbon loading) systems. As such, the N₂:Ar flux technique represents a more realistic estimation of net denitrification in sediments with low organic carbon contents and low respiration rates and significant bioturbation. Most importantly, the present study also demonstrated that methodological comparisons need to be made over a range of environmental conditions.