Background: Volatile anaesthetics are widely used in human medicine. Although generally safe, hypersensitivity and toxicity can occur in rare cases, such as in certain genetic disorders. Anaesthesia hypersensitivity is well-documented in a subset of mitochondrial diseases, but whether volatile anaesthetics are toxic in this setting has not been explored. Methods: We exposed Ndufs4(−/−) mice, a model of Leigh syndrome, to isoflurane (0.2–0.6%), oxygen 100%, or air. Cardiorespiratory function, weight, blood metabolites, and survival were assessed. We exposed post-symptom onset and pre-symptom onset animals and animals treated with the macrophage depleting drug PLX3397/pexidartinib to define the role of overt neuroinflammation in volatile anaesthetic toxicities. Results: Isoflurane induced hyperlactataemia, weight loss, and mortality in a concentration- and duration-dependent manner from 0.2% to 0.6% compared with carrier gas (O 2 100%) or mock (air) exposures (lifespan after 30-min exposures ∗P<0.05 for isoflurane 0.4% vs air or vs O 2, ∗∗P<0.005 for isoflurane 0.6% vs air or O 2; 60-min exposures ∗∗P<0.005 for isoflurane 0.2% vs air, ∗P<0.05 for isoflurane 0.2% vs O 2). Isoflurane toxicity was significantly reduced in Ndufs4(−/−) exposed before CNS disease onset, and the macrophage depleting drug pexidartinib attenuated sequelae of isoflurane toxicity (survival ∗∗∗P=0.0008 isoflurane 0.4% vs pexidartinib plus isoflurane 0.4%). Finally, the laboratory animal standard of care of 100% O 2 as a carrier gas contributed significantly to weight loss and reduced survival, but not to metabolic changes, and increased acute mortality. Conclusions: Isoflurane is toxic in the Ndufs4(−/−) model of Leigh syndrome. Toxic effects are dependent on the status of underlying neurologic disease, largely prevented by the CSF1R inhibitor pexidartinib, and influenced by oxygen concentration in the carrier gas.