Objective: To explore the reliability of acute physiology and chronic health evaluation II (APACHE II) to evaluate the severity of the neurologic diseases and its accuracy to predict the outcome of patients with these diseases. Methods: Four hundred and four consecutive patients with severe neurologic diseases between 2005 and 2006 were enrolled to obtain the APACHE II scores at 0, 24, 48, 72 h after admission to neurointensive care unit. Results: The APACHE II scores were positively associated with the outcome of the patients with severe neurologic diseases. The higher score corresponded with the higher mortality rate. The areas under the receiver operating characteristic curves of APACHE II to predict the outcome was 0.866(95% CI 0.824 to 0.907, P = 0.000). The optimal cutoff of APACHE II scores to predict the outcome was 17 scores with the sensitivity of 76.7% and the specificity of 78.7%. The predictive chance that was mostly associated with the outcome was 72 h after admission in logistic regression model (X 2 = 137.345, P = 0.000, correct class% = 85%). The factors that were the most statistically associated with the outcome in the 14 parameters of APACHE II were GCS score, heart rate, serum creatinine, body temperature and WBC count Conclusion: APACHE II favorably reflects the severity of the neurologic diseases and reliably and accurately predicts the prognosis of the patients.