Although the C-reactive protein (CRP) concentration measured shortly after acute myocardial infarction (AMI) is associated with infarct size, its prognostic value is controversial. The reduction of CRP is accelerated by reperfusion. Therefore, the CRP concentration, measured during the stable phase of AMI in patients treated predominantly with reperfusion therapies, may be independent of infarct size and may predict long-term mortality. We studied 1,309 patients with AMI enrolled in the Osaka Acute Coronary Insufficiency Study between April 1999 and June 2001. CRP was measured during the stable phase (mean 25 days after AMI onset). The patients were followed for an average of 522 days. Reperfusion therapies were performed in 90% of the patients. Patients in the highest quartile of CRP values (> or =0.38 mg/dl) were older, had higher prevalences of diabetes mellitus, and had higher Killip classes than patients in the lower 3 quartiles (<0.38 mg/dl). Multivariate logistic regression analysis revealed that CRP was independently associated with age and the absence of revascularization therapies. Patients in the highest quartile had a higher long-term mortality rate than patients in the lower 3 quartiles (8.9% vs 2.0%; p <0.001). Multivariate Cox regression analysis revealed that the highest quartile of CRP values was an independent predictor of long-term mortality (hazard ratio 4.94, 95% confidence interval 1.13 to 21.6). We conclude that CRP measured during the stable phase of AMI is not associated with infarct size in the reperfusion era but is significantly associated with long-term mortality of AMI.