An inverse relation exists between intake of flavonoid-rich foods, such as cocoa, and cardiovascular-related mortality. Favorable effects of flavonoids on the endothelium may underlie these associations. We performed a randomized, double-blind, placebo-controlled study to test the hypothesis that acute cocoa ingestion dose dependently increases endothelium-dependent vasodilation, as measured by an increase in brachial artery flow-mediated dilation (FMD), in healthy older adults. Measurements were obtained before (preingestion) and after (1- and 2-h postingestion) ingestion of 0 (placebo), 2, 5, 13, and 26 g of cocoa in 23 adults (63 ± 2 yr old, mean ± SE). Changes in brachial artery FMD 1- and 2-h postingestion compared with preingestion were used to determine the effects of cocoa. FMD was unchanged 1 (Δ-0.3 ± 0.2%)- and 2-h (Δ0.1 ± 0.1%) after placebo (0 g cocoa). In contrast, FMD increased both 1-h postingestion (2 g cocoa Δ0.0 ± 0.2%, 5 g cocoa Δ0.8 ± 0.3%, 13 g cocoa Δ1.0 ± 0.3%, and 26 g cocoa Δ1.6 ± 0.3%: P < 0.05 compared with placebo for 5, 13, and 26 g cocoa) and 2-h postingestion (2 g cocoa Δ0.5 ± 0.3%, 5 g cocoa Δ1.0 ± 0.3%, 13 g cocoa Δ1.4 ± 0.2%, and 26 g cocoa Δ2.5 ± 0.4%: P < 0.05 compared with placebo for 5, 13, and 26 g cocoa) on the other study days. A serum marker of cocoa ingestion (total epicatechin) correlated with increased FMD 1- and 2-h postingestion (r = 0.44-0.48; both P < 0.05). Collectively, these results indicate that acute cocoa ingestion dose dependently increases brachial artery FMD in healthy older humans. These responses may help to explain associations between flavonoid intake and cardiovascular-related mortality in humans.