Objective To assess the predictive value of serum uric acid levels for cardiovascular and all-cause mortality in a large prospective population based study. Methods The study was based on 3648 participants in Shanghai and Beijing, who were inpatients with high cardiovascular(CV) risk at baseline (2004 .7 to 2005.1), and blood was taken. Follow-up for death from cardiovascular disease and any cause was complete until January 1, 2006. Results The mean follow-up was 1 years. There were 303 deaths during follow-up, of which 121 were cardiovascular. Crude mortality rates were 8.3 % for all patients, 6.8% for female patients (116/1715), and 9.7% (187/ 1933) for male patients. Among men, patients in the lower and higher uric acid groups had increased cardiac and overall mortality risks compared with patients in the normal uric acid groups. Similar relation was found in women but not statistically significant. After adjusting for other conventional risk factors (age, diabetes, hypertension, diuretic use and smoking), baseline uric acid level was still associated with increased risk for death from cardiovascular disease (P=0.005), or death from all causes (P=0.014) Conclusion Our data suggest that abnormal serum uric acid levels are independently and significantly associated with risk of cardiovascular and all-causemortality. Methods The study was based on 3648 participants in Shanghai and Beijing, who were inpatients with high cardiovascular (CV) risk at baseline (2004 .7 to 2005.1), and blood was taken. Follow-up for death from cardiovascular disease and any cause was complete until January 1, 2006. There were 303 deaths during follow- Crude mortality rates were 8.3% for all patients, 6.8% for female patients (116/1715), and 9.7% (187/1933) for male patients. Among men, patients in the lower and higher uric acid groups had increased cardiac and overall mortality risks compared with patients in the normal uric acid groups. Similar adjusted was other women risk factors (age, diabetes, hypertension, di uretic use and smoking), baseline uric acid level was still associated with increased risk for death from cardiovascular disease (P = 0.005), or death from all causes (P = 0.014) Conclusion Our data suggest that abnormal serum uric acid levels are independently and significantly associated with risk of cardiovascular and all-causemortality.