OBJECTIVES: Hyperbaric oxygen therapy (HBO2) is recommended for symptomatic patients within 24-hour postcarbon monoxide poisoning. Previous studies have reported significantly better outcomes with treatment administered within 6 hours after carbon monoxide poisoning. Thus, we aimed to compare the neurocognitive outcomes according to HBO2 delay intervals. DESIGN: Retrospective analysis of data from our prospectively collected carbon monoxide poisoning registry. SETTING: A single academic medical center in Wonju, Republic of Korea. PATIENTS: We analyzed the data of 706 patients older than 16 years treated with HBO2 with propensity score matching. Based on carbon monoxide exposure-to-HBO2 delay intervals, we classified patients into the early (control, less than or equal to 6 hr) and late (case, 6-24 hr) groups. The late group was further divided into Case-1 (6-12 hr) and Case-2 (12-24 hr) groups. We also compared mild (nonintubated) and severe (intubated) groups. INTERVENTIONS: HBO2. MEASUREMENTS AND MAIN RESULTS: After propensity score matching, Global Deterioration Scale scores at 6 months postcarbon monoxide exposure showed significantly fewer poor outcome patients in the early than in the late group (p = 0.027). The early group had significantly fewer patients with poor outcomes than the Case-2 group (p = 0.035) at 1 month and than the Case-1 (p = 0.033) and Case-2 (p = 0.004) groups at 6 months. There were significantly more patients with poor prognoses at 6 months as treatment interval increased (p = 0.008). In the mild cohort, the early group had significantly fewer patients with poor 6-month outcomes than the late group (p = 0.033). CONCLUSIONS: Patients who received HBO2 within 6 hours of carbon monoxide exposure had a better 6-month neurocognitive prognosis than those treated within 6-24 hours. An increase in the interval to treatment led to an increase in poor outcomes.