The mechanism of risk reduction obtained by blood pressure-lowering pharmacological treatment remains unclear. We explored the amount of risk reduction attributable to the apparent effect of antihypertensive medicines on blood pressure by using the capture approach. Five randomized, placebo or nil controlled trials with a total of 28,997 subjects and 1,935 cardiovascular fatal or non-fatal events from the INDANA database met the eligibility criteria. Computations were performed on the original individual records using multiple Cox's proportional hazard regression models designed for meeting the assumed treatment mode of action and comparing relevant assumptions. For coronary event, the results are inconclusive essentially because the risk reduction is mild. However, for stroke the risk reduction adjusted for baseline risk factors is 34% (P<0.0001). The part explained by the effect of treatment on systolic blood pressure is 49% of this reduction, with 95% confidence interval not including 100%. This result suggests that the apparent effect on blood pressure is not the only cause of stroke risk reduction in hypertensive subjects submitted to an antihypertensive medicine.