Bradykinin stimulates tissue plasminogen activator release from human endothelium through a flow-independent, B2 receptor-dependent mechanism. The present study tests the hypothesis that smoking impairs bradykinin-stimulated tissue plasminogen activator release. Graded doses of nitroprusside (1.6 to 6.4 microg/min), methacholine (3.2 to 12.8 microg/min), and bradykinin (100 to 400 ng/min) were infused in the brachial artery in random order in 20 smokers and 12 nonsmokers matched for age, gender, and body mass index. Forearm blood flow was measured by strain-gauge plethysmography. All 3 drugs caused a dose-dependent increase in forearm blood flow, with no significant difference between smokers and nonsmokers. Bradykinin (P=0.001) and methacholine (P=0.001) caused significant dose-dependent increases in net tissue plasminogen activator release. The tissue plasminogen activator response to bradykinin was significantly greater than the tissue plasminogen activator response to methacholine in the nonsmokers (maximal net tissue plasminogen activator release, 73.2+/-21.5 versus 27.6+/-7.2 ng/min per 100 mL; P=0.001) but not in the smokers (maximal net tissue plasminogen activator release, 44.5+/-10.7 versus 24.8+/-9.3 ng/min per 100 mL; P=0.154). The effect of bradykinin (P=0.037), but not methacholine (P=0.978), on net tissue plasminogen activator release was significantly reduced in smokers compared with nonsmokers. The vascular tissue plasminogen activator response to bradykinin, but not methacholine, is impaired in smokers. Stimulated tissue plasminogen activator release may be a more sensitive measure of endothelial function than vasodilation.