A 59-year-old man undergoing a major depressive episode used charcoal briquettes in X year. The following day, the patient was transported to the emergency department and was diagnosed with CO poisoning with loss of consciousness. HBO therapy was administered seven times within 2 weeks, and the patient was discharged thereafter as no depressive symptoms were observed. Two weeks after discharge, the patient presented with memory impairment and difficulty in bathing. He was admitted to our hospital 1 month after CO exposure and was diagnosed with intermittent CO poisoning. He had a Hasegawa Dementia Scale-Revised (HDS-R) (a brief cognitive scale) score of 16 points (the cut-off point for dementia is 21/20). HBO therapy was resumed but his cognitive function declined further (HDS-R score 2 points). Eventually, he developed akinetic mutism. After the 30th HBO therapy session, his cognitive function recovered slightly (HDS-R score 9 points), but he gradually became agitated and violent, therefore drug therapy was initiated. Valproic acid (800 mg/day), risperidone (1 mg/day), and quetiapine (50 mg/day) were prescribed, resulting in the alleviation of the patient's agitation. After the 65th HBO therapy session, his HDS-R score increased to 22 points, and the patient was discharged after 99 days of hospitalization. HBO therapy was continued in the outpatient setting. After the 89th HBO therapy session, the HDS-R score increased to 29 points (out of 30), and HBO therapy was terminated. The patient continued to take valproic acid only from X ＋ 1 years after discharge. Neuroimaging results are shown in Figures,,,. Magnetic resonance imaging (MRI) revealed deep white matter hyperintensities around the anterior and posterior horns of the lateral ventricles during HBO therapy. At 2 and 8 years after discharge, MRI showed a gradual spread of deep white matter hyperintensity and overall brain atrophy. No changes were observed in basal ganglia, including globus pallidus. In contrast, single-photon emission computed tomography (SPECT) showed improvement in the extensive frontal hypoperfusion and in the basal ganglia, hippocampus, and periventricular areas of the lateral ventricles during HBO therapy. At 8 and 9 years after discharge, SPECT showed decreased blood flow, mainly in the frontoparietal lobe. However, it was not as extensive as the baseline image. There was no apparent re-decrease in blood flow in the basal ganglia or other areas. The patient returned to work as an engineer 3 months after his discharge in X year and became a director in X + 3 years, then retired at the age of 65 years, but was rehired by the same employer. In X + 8 years, the patient had an HDS-R score of 30 points (full score), a frontal assessment battery score of 15 points (frontal lobe dysfunction cutoff point of 12), and a full intelligence quotient (IQ), measured using the Wechsler Adult Intelligence Scale-III, of 107 (verbal IQ 104, performance IQ 109). These scores indicated no cognitive dysfunction. Additionally, no apparent decline in WAIS-3rd index scores (verbal comprehension 102 points, perceptual organization 110 points, working memory 113 points, processing speed 113 points) was observed. Treatment with valproic acid was stopped in X + 9 years. From then until X + 10 years, treatment was terminated, as the patient was able to work without medication.