We sought to test the hypothesis of an antioxidant effect of dipyridamole in vivo in a model of cerebral hypoperfusion. Twenty-one patients (65+/-10 years, 11 men) undergoing carotid endarterectomy were allocated in two groups (group 1, 10 with dipyridamole, 200 mg p.o., 3-4 h before surgery; group II, 11 with placebo) in a double-blind placebo-controlled randomized design. Blood was sampled from ipsilateral jugular bulb, and plasma vitamin E content was assayed before, after 15 and 30 min of clamp, and 2 and 10 min after declamping. In 12 of them, lipoperoxides were assayed. Vitamin E plasma content decreased significantly in group II (rest, 3.71+/-0.22 mmol/mol of cholesterol, 100%) after clamp (91.5% of rest, p < 0.01) and remained unchanged during declamping (90.9% of rest, p < 0.01), but did not change in group I (rest, 3.5+/-0.44 mmol/mol of cholesterol, 100%) during clamping (99.9% of rest; p = NS) and after declamping (97.6% of rest; p = NS). Lipoperoxide concentration did not change in group I (rest, 302+/-8 a.u.; clamp, 296+/-13 a.u.; p = NS vs. rest; declamp, 304+/-8 a.u.; p = NS vs. rest), and increased significantly in group II (rest, 313+/-5 a.u.; clamp, 352+/-9 a.u.; p < 0.01 vs. rest; declamp, 343+/-6 a.u.; p < 0.05 vs. rest). Cerebral oxidative stress associated with human carotid endarterectomy can be attenuated by pretreatment with oral dipyridamole.