Metabolic rate depression (MRD) has long been proposed as the key metabolic strategy of hypoxic survival, but surprisingly, the effects of changes in hypoxic O 2 tensions (Pw O2 ) on MRD are largely unexplored. We simultaneously measured the O 2 consumption rate (Ṁ O2 ) and metabolic heat of goldfish using calorespirometry to test the hypothesis that MRD is employed at hypoxic Pw O2 values and initiated just below P crit , the Pw O2 below which Ṁ O2 is forced to progressively decline as the fish oxyconforms to decreasing Pw O2 . Specifically, we used closed-chamber and flow-through calorespirometry together with terminal sampling experiments to examine the effects of Pw O2 and time on Ṁ O2 , metabolic heat and anaerobic metabolism (lactate and ethanol production). The closedchamber and flow-through experiments yielded slightly different results. Under closed-chamber conditions with a continually decreasing Pw O2 , goldfish showed a P crit of 3.0±0.3 kPa and metabolic heat production was only depressed at Pw O2 between 0 and 0.67 kPa. Under flow-through conditions with Pw O2 held at a variety of oxygen tensions for 1 and 4 h, goldfish also initiated MRD between 0 and 0.67 kPa but maintained Ṁ O2 to 0.67 kPa, indicating that P crit is at or below this Pw O2 . Anaerobic metabolism was strongly activated at Pw O2 ≤1.3 kPa, but only used within the first hour at 1.3 and 0.67 kPa, as anaerobic end-products did not accumulate between 1 and 4 h exposure. Taken together, it appears that goldfish reserve MRD for near-anoxia, supporting routine metabolic rate at sub-P crit Pw O2 values with the help of anaerobic glycolysis in the closed-chamber experiments, and aerobically after an initial (<1 h) activation of anaerobic metabolism in the flow-through experiments, even at 0.67 kPa Pw O2 .