OBJECTIVES: To evaluate the Full Outline of Unresponsiveness (FOUR) score as a predictor of outcome in adult patients with bacterial meningitis. METHODS: We selected 427 patients from a nationwide, prospective cohort on community-acquired bacterial meningitis included from August 2011 to November 2016. Data on patient history, symptoms and signs on admission, treatment, and outcome were collected. We compare the FOUR score with the Glasgow Coma Scale (GCS) score, performed a receiver operator characteristic curve analysis, and calculated the area under the curve (AUC) of the FOUR and GCS scores for the prediction of unfavorable outcome and mortality. RESULTS: The median FOUR score on admission was 14 (interquartile range [IQR] 12-16), and the median GCS score was 12 (IQR 9-14). The outcome was unfavorable in 135 of 427 (32%) patients, of whom 55 (13%) died. There was a strong correlation between the FOUR score and the GCS score (r = 0.85, p < 0.001). AUCs for the GCS and FOUR scores in the prediction of unfavorable outcome (both 0.64) and mortality (both 0.68) were comparable. Logistic regression analysis showed that the FOUR motor, brainstem, and respiration items were individual predictors of unfavorable outcome and mortality. For the GCS score, only the motor component was predictive, while the FOUR score provided a spectrum of clinical abnormalities in patients with a GCS score of 3. CONCLUSIONS: The FOUR score adds considerably to the prediction of outcome in patients with severe meningitis by means of better testing of the brainstem reflexes and respiration status. Future studies should consider incorporating the FOUR score into clinical assessment.