Objective: Survey serum S100 and NSE concentration in patients with cerebral infarction during the acute phase. To understand the relationship between serum S100 and NSE concentrations with some other risk factors such as age, sex, Glasgow Coma Scale, cerebral lesion volume on computerized tomography. Subjects and Methods: Study of 90 hospitalized patients with cerebral infarction at ICU of Hue Central Hospital  and 100 controls. Data was collected through medical records of patients with acute cerebral infarction and control. The study method was acrossectional and descriptive. Data was analysed  by medical statistics and processed by the SPSS 19.0 software. Results: The average age in study group was 68.3 ± 13.1 (Min 32, Max 90) and control group was 64.8 ± 12.9 (Min 33, Max 88) did not differ statistics significantly. The majority of patients with age group from 61 to 80 years old (48.9%) is retired elderly patients but it doesn’t differ statistics significantly with p> 0.05. The average concentration of S100 and NSE in study group was higher in control group, the difference was statistics significantly(p<0.001). The concentration of  S100 and NSE in mortality was higher in living groups, in which S100 was different statistics significantly (p <0.05). Cut-off value  predicts the survival of S100 and NSE in this study is respectively 0.21 mcg/l and 20.45ng/ml. There is a inversely correlation between the S100 and NSE with Glasgow coma scale. The more comatose patients are, the more NSE and S100 increases. The correlation equation respectively: y = -0.1975x10.439 + (n = 90, r= - 0.19, p> 0.05). Y = -0.0228x + 11.02 (n = 90, r = - 0.29, p <0.01). The greater volume of lesions is, the higher the concentration  of  S100 and NSE is. There is a agreement correlation between the S100 and NSE with lesion volume, the correlation equations respectively: y = 20.6x + 67.71 (n = 90, r = 0.397, p <0.001). Y = 1.441x + 43.104 (n = 90, r = 0.359, p <0.05). Conclusion: The cut-off value to predict the survival of S100 and NSE in this study respectively 0.21 mcg/l and 20.45ng/ml.There is a inversely correlation between S100 and NSE with Glasgow coma scale. There is positive correlation between S100 and NSE with lesion volume. S100 and NSE can be used to predict and monitor disease  progression and the volume of cerebral lesions.