PURPOSE: The Framingham risk score is a popular tool for estimating cardiovascular risk, but there is debate about the value of C-reactive protein. This study investigated the utility of C-reactive protein to predict the progression of atherosclerosis in relation to the Framingham risk score and age. METHODS: This observational study enrolled 164 outpatients (mean age: 61 yr; range: 40 - 75 yr) receiving treatment for classical cardiovascular risk factors. They underwent serial ultrasonographic evaluation of their carotid arteries for 36 +/- 10 months. A carotid intima-media thickness > or = 1.1 mm was defined as plaque, and the number of plaques and plaque score (sum of all plaque thicknesses) were determined. Serum C-reactive protein concentrations and classical risk factors, including body mass index, were measured. RESULTS: C-reactive protein was related to annual changes in the number of plaques and the plaque score (r=0.26 and 0.28; P<0.01 and P<0.001, respectively), as well as the 10-year risk of cardiovascular disease estimated from the Framingham risk score (r=0.335, P<0.001). C-reactive protein was correlated with the annual changes of plaque number and plaque score (beta=0.21 and 0.23; P<0.05 and P<0.01) after adjusting for 10-year cardiovascular and other risk factors, especially in 64 patients comprising the 8-13% 10-year risk group (beta=0.33 for plaque score, P<0.05). CRP also showed a relationship with the progression of carotid atherosclerosis in 71 patients aged < or = 61 yr (beta=0.33 for plaque score, P<0.01). CONCLUSIONS: C-reactive protein can predict the progression of early carotid atherosclerosis in patients with mild to moderate cardiovascular risk and/or middle-aged patients.