Background: Epidemiologic studies have suggested that flavonoid intake plays a critical role in the prevention of coronary heart disease. Because atherosclerosis is considered a low-grade inflammatory disease, some feeding trials have analyzed the effects of cocoa (an important source of flavonoids) on inflammatory biomarkers, but the results have been controversial. Objective: The objective was to evaluate the effects of chronic cocoa consumption on cellular and serum biomarkers related to atherosclerosis in high-risk patients. Design: Forty-two high-risk volunteers (19 men and 23 women; mean 6 SD age: 69.7 6 11.5 y) were included in a randomized crossover feeding trial. All subjects received 40 g cocoa powder with 500 mL skim milk/d (C+M) or only 500 mL skim milk/d (M) for 4 wk. Before and after each intervention period, cellular and serum inflammatory biomarkers related to atherosclerosis were evaluated. Results: Adherence to the dietary protocol was excellent. No significant changes in the expression of adhesion molecules on T lymphocyte surfaces were found between the C+M and M groups. However, in monocytes, the expression of VLA-4, CD40, and CD36 was significantly lower (P = 0.005, 0.028, and 0.001, respectively) after C+M intake than after M intake. In addition, serum concentrations of the soluble endothelium-derived adhesion molecules P-selectin and intercellular adhesion molecule-1 were significantly lower (both P = 0.007) after C+M intake than after M intake. Conclusions: These results suggest that the intake of cocoa polyphenols may modulate inflammatory mediators in patients at high risk of cardiovascular disease. These antiinflammatory effects may contribute to the overall benefits of cocoa consumption against atherosclerosis. This trial was registered in the Current Controlled Trials at London, International Standard Randomized Controlled Trial Number, at controlled-trials.com as ISRCTN75176807.