The evaluation of antithrombotic agents for secondary stroke prevention has focused on stroke reduction. The aim of this analysis was to focus specifically on the increase in bleeding risk. The annualized rates of total and major bleeding events in secondary stroke prevention trials of antithrombotics were assessed and cross compared. A Medline search for major randomized clinical studies with a follow-up duration of > or =1 year identified 13 studies. Pooled data sets were used to compare mean bleeding rates for aspirin (< or =325 mg/day), clopidogrel, anticoagulants (warfarin and other vitamin K antagonists), aspirin plus clopidogrel, and aspirin plus extended-release dipyridamole (ER-DP). Total bleeding occurred at mean rates of 4.8% with aspirin (< or =325 mg/day) alone, 2.9% with clopidogrel alone, 3.6% with aspirin plus ER-DP, 10.1% with aspirin plus clopidogrel, and 16.8% with anticoagulation. Major bleeding occurred at mean rates of 1% with aspirin (< or =325 mg/day) alone, 0.85% with clopidogrel, 0.93% with aspirin plus ER-DP, 1.7% with aspirin plus clopidogrel, and 2.5% with anticoagulation. In conclusion, the combination of aspirin and clopidogrel is associated with significantly greater bleeding than either aspirin (< or =325 mg/day) or clopidogrel alone. Aspirin plus ER-DP has a greater bleeding rate than clopidogrel but a lower rate than aspirin (< or =325 mg/day) alone.