ABSTRACT Understanding the critical thresholds of dissolved oxygen (O2) that trigger adaptive physiological responses in aquatic organisms is long hampered by a lack of robust, non-lethal or non-invasive methodologies. The isotope fractionation of triple O2 isotopes (18O/17O/16O) during respiration is linked to the amount of oxygen utilised, offering a potential avenue for new insights. Our experimental research involved measuring the oxygen isotope fractionation of dissolved O2 in closed-system aquatic respirometry experiments with wild sticklebacks (Gasterosteus aculeatus). These fish were either naturally adapted or experimentally acclimated to hypoxic and normoxic conditions. The aim was to observe their oxygen usage and isotope fractionation in response to increasingly severe hypoxia. Initial observations revealed a progressive 18O enrichment from the preferential uptake of 16O to a dissolved oxygen threshold of 3–5 mg O2 L–1, followed by an apparent reversal in oxygen isotope fractionation, which is mixing of 16O and 17O with the remaining O2 pool across all populations and indicative of a systematic change in oxygen metabolism among the fish. Unexpectedly, sticklebacks adapted to hypoxia but acclimated to normoxia exhibited stronger oxygen isotope fractionation compared to those adapted to normoxia and acclimated to hypoxia, contradicting the hypothesis that hypoxia adaptation would lead to reduced isotope discrimination due to more efficient oxygen uptake. These preliminary experimental results highlight the novel potential of using dissolved O2 isotopes as a non-invasive, non-lethal method to quantitatively assess metabolic thresholds in aquatic organisms. This approach could significantly improve our understanding of the critical oxygen responses and adaptation mechanisms in fish and other aquatic organisms across different oxygen environments, marking a significant step forward in aquatic ecological and physiological research.