Previous aggregation studies in healthy volunteers have demonstrated contrasting results when platelets suspended in buffer or plasma are exposed to reactive oxygen species (ROS). We investigated the effect of ROS on whole blood aggregation (WBA) and the endothelial cell-platelet interaction in coronary heart disease (CHD) patients with and without diabetes. Data are expressed as mean difference with 95% confidence intervals. 14 patients with CHD (7 with diabetes) were recruited to study the effects of ROS on WBA. ROS generated by xanthine (X) and xanthine oxidase (XO) caused a significant dose dependent inhibition of WBA. 100 μM X and 100 mU/ml XO significantly inhibited WBA in response to 3 μg/ml collagen (28.9%, CI 16.8%-40.9%, p<0.01) and 5 μM ADP (36.0%, CI 13.5%-58.5%, p<0.01). Results were similar for both diabetic and non-diabetic subgroups. An additional 9 patients with CHD were recruited for endothelial cell-platelet interaction studies. Addition of 1 x 10⁵ endothelial cells cultured from porcine pulmonary artery caused a significant reduction in WBA in response to 3 μg/ml collagen (31.2%, CI 11.0%-51.5%, p<0.01) and 5 μM ADP (31.6%, CI 0.5-62.6%, p<0.05). Exposing the endothelial cells to 1 minute of ROS using 100 μM X and 10 mU/ml XO did not attenuate the inhibitory effect. In 5 patients, endothelial cells were pretreated with 100 μM L-NAME and this did not alter the effect of endothelial cells significantly. In conclusion, the acute dose dependent effect of ROS on WBA is inhibitory and may differ from the effects of ROS on platelets in less physiological environments such as buffer or plasma. ECs inhibit WBA in patients with CHD. The anti-aggregatory effect of ECs is not prevented by addition of exogenous ROS or L-NAME pretreatment, probably because it involves a predominant non-nitric oxide mechanism.