BACKGROUND: Serum C-reactive protein (CRP) has proven to be an independent marker of the extent of atherosclerosis in patients with coronary, cerebrovascular, and peripheral arterial disease. In this prospective observational study, we wanted to assess the relationship between serum CRP and extent of disease transversely and longitudinally in time, as well as future cardiovascular complications in patients with peripheral arterial disease (PAD). Hypothesizing that CRP not only is a marker of but also actively participates in atherogenesis, we explored the possibility of CRP production by femoral atherosclerotic plaques. METHODS: Serum CRP was measured as highly sensitive (hsCRP) in 387 patients with PAD attending the vascular clinic of a university and 2 affiliated teaching hospitals. Serum hsCRP was related to the ankle-brachial pressure index (ABPI) as an indication of severity of disease at inclusion and at 12 months' follow-up and to future events (death and coronary, cerebral, and peripheral arterial events). In femoral plaques, the production of CRP was analyzed with reverse transcription-polymerase chain reaction, and CRP plaque localization was assessed with immunostaining on serial tissue sections with antibodies toward CRP, smooth muscle cells, T cells, and macrophages. RESULTS: The hsCRP (average +/- SD) was 3.26 +/- 2.41 mg/L. Serum hsCRP showed a correlation with baseline and 12-month follow-up ABPI (Spearman rank correlation; P < .05 for both correlations). When the patients were divided into three equally sized groups according to baseline serum hsCRP, the ABPI at baseline and at 12 months decreased significantly from the low- to the high-hsCRP group (baseline ABPI: 0.70, 0.65, and 0.57, P < .01; 12-month follow-up ABPI: 0.78, 0.70, and 0.65, P < .01). These associations persisted after correction for conventional risk factors. Furthermore, serum hsCRP was related to the combined end point "death and/or any cardiovascular event" (log-rank test; P = .04) during a median 24-month follow-up period. Reverse transcription-polymerase chain reaction analysis showed CRP production in 4 of 14 femoral plaques. CRP was detected in all femoral plaques, but not in healthy brachial arteries. Immunoreactivity for CRP was observed in smooth muscle cells, macrophages, and T cells. CONCLUSIONS: Serum hsCRP was related to the severity of PAD, showing a relation to future hemodynamic function and cardiovascular events in PAD patients. In addition to coronary plaques, aneurysmal aortas, and failed venous coronary bypasses, femoral plaques also produce CRP, thus illustrating that the production of CRP may represent a universal response to vascular injury and suggesting that vascular CRP may contribute to plaque development.