Background: Hyponatremia is the most common electrolyte disorder in hospitalized patients, and is frequently a marker of a significant underlying disease. The prognostic value of hyponatremia in patients with acute first-ever ischemic stroke is not known. We aimed to analyze whether hyponatremia in the acute stroke stage contributed to the risk of mortality or recurrent stroke in these patients. Methods: We studied 925 patients presenting with acute first-ever ischemic stroke between 2002 and 2004. Sodium levels were obtained on arrival at the emergency room within 3 days of acute stroke onset. Hyponatremia was defined as a serum sodium concentration of 134 mmol/l or less. Clinical presentation, stroke risk factors, associated medical disease, and outcome were recorded. All patients were followed for 3 years for survival analysis. A multivariate Cox proportional hazards model was used to identify risk factors for 3-year mortality in these patients. We also constructed Kaplan-Meier survival curves, and compared groups with hyponatremia and normonatremia by means of log rank tests for significant differences. Results: Among the patients with acute first-ever ischemic stroke, 107 (11.6%) were hyponatremic. Among stroke risk factors, the prevalence of diabetes mellitus was significantly higher among hyponatremic patients (p < 0.001). Prevalence of chronic renal insufficiency was also higher in the hyponatremic group (p = 0.002). Clinical presentations, such as the length of acute ward stay, initial impaired consciousness, and clinical course in acute stroke were similar among normo- and hyponatremic patients. Among the complications, pneumonia and urinary tract infection were significantly higher in hyponatremic than in normonatremic patients. After multivariate logistic regression analysis, diabetes mellitus and chronic renal insufficiency were associated with hyponatremia in these patients. Kaplan-Meier analysis indicated that the survival rate was significantly lower in hyponatremic patients than in normonatremic patients (log rank test; p value <0.001). After multivariate Cox proportional hazards model analysis, hyponatremia was a significant predictor of 3-year mortality in these patients after adjustment for related variables (p value = 0.003, hazard ratio = 2.23, 95% confidence interval: 1.30–3.82). Conclusion: Hyponatremia in the acute stroke stage is a predictor of 3-year mortality in patients with acute first-ever ischemic stroke that is independent of other clinical predictors of adverse outcome.