Hypoxia is thought to downregulate ENaC in alveolar epithelia and has been implicated in the formation of and failure to clear pulmonary oedema at high altitude. However the mechanisms involved remain unclear. Channel trafficking, direct channel modification and altered subunit transcription have all been implicated, but the evidence is conflicting. Recently, it has been reported that in 3T3 cells stably transfected with ENaC, channel activity ran down when the cells were dialysed with an ATP-free solution (Ishikawa et al. 2003). Here we examined the effect of hypoxia on rat ENaC heterologously expressed in Xenopus laevis oocytes, and sought to see whether this was related to a change in ATP levels using the two-electrode voltage clamp technique, a surface expression and a luciferase-based ATP assay. Data are given as means ± S.E.M.; significance was evaluated using the appropriate Student’s t tests. Following 18 hours of continuous exposure to anoxia, amiloride-sensitive currents (ΔI_Ami) in ENaC-expressing oocytes averaged 0.2 ± 0.0 μA (n = 83) compared with 5.7 ± 0.7 μA in time-matched normoxic oocytes (n = 74; p < 0.001). The effect of anoxia was reversible and after 6 hours of re-exposure to normoxia, ΔI_Ami averaged 5 ± 0.5 μA (n = 64). Despite a reduction in ΔI_Ami by 85.1 ± 2.0 % (p < 0.001) in anoxic (n = 20) compared to normoxic oocytes (n = 19), channel surface expression did not differ between the two groups (n = 37 and 35, respectively). Measurements of cytosolic ATP in normoxic oocytes averaged 2.4 ± 0.5 mM, compared with 0.9 ± 0.2 mM in anoxic oocytes (n = 6 each; p < 0.05). Injection of exogenous ATP in anoxic oocytes resulted in a partial recovery of ΔI_Ami within three hours; in these oocytes ΔI_Ami averaged 51.7 ± 4.4 % of that seen in anoxic oocytes re-exposed to normoxia. Using a Na⁺-free perfusate to prevent rundown of channel activity as described by Volk et al. (2004), continuous current recordings were made over a 30-minute period. When the metabolic inhibitor sodium azide was added to the perfusate, we observed a 77.6 ± 4.7 % rundown in ΔI_Ami (n = 21; p < 0.001). There was no change in ENaC surface expression following incubation with sodium azide, thus mirroring the effects seen with anoxia. Together these data suggest that [ATP]_i is essential to maintain ENaC activity and that the reduction in [ATP]_i during anoxia is responsible for the decrease in ENaC activity.