Background: Perioperative cardiac complication is a major cause of morbidity after noncardiac surgery. We hypothesized that measurement of cardiovascular biomarkers may have an additional value in predicting perioperative cardiac complications. Methods: From 2004 to 2007, 2054 patients referred for cardiac consultation before elective noncardiac surgery was enrolled prospectively. Revised cardiac risk index (RCRI), which is currently regarded as the standard prediction methods, was obtained and HsCRP and NT-proBNP were measured within 2 weeks before surgery. Primary endpoint was major cardiaovascular event (MCE) including all-cause death, myocardial infarction (MI), aggravation of heart failure, and revascularization within 30 days. Results: MCEs occurred in 290 (14.12%) patients, which included 102 (4.97%) MI, 248 (12.1%) heart failure, 20 (0.97%) deaths, and 26 (1.27%) revascularization procedures. The optimal cut-off value of NT-proBNP (194 pg/ml, sensitivity 70.34%, specificity 70.18%) and CRP (0.335 mg/dL, sensitivity 66.79%, specificity 66.42%) was derived from logistic regression and used for subsequent analysis. Both higher preoperative NT-proBNP (OR=4.36, 95% CI=3.24-5.89) and hsCRP (OR=3.02, 95% CI=2.26-4.05) as well as higher RCRI (OR=1.40, 95% CI = 1.18 – 1.66) were independent predictor of MCEs (p <0.01). Next, we compared the performance of predicting POCEs by comparing the ROC curves. Both NT-proBNP (AUC=0.753) and CRP (AUC=0.723) showed better prediction power than RCRI (AUC=0.627, p<0.001 for RCRI versus NT-proBNP or CRP). Furthermore, integration of these biomarkers into RCRI resulted in much better predictive method (AUC=0.794, p < 0.05). Conclusions: High preoperative NT-proBNP (��� 194 pg/ml) or hsCRP (��� 0.335 mg/dl) is a strong predictor of perioperative major cardiac events upto 30 days. Measurement of cardiovascular biomarkers can significantly strengthen the current standard risk prediction methods.