Although we recognize, along with Davey Smith and Ebrahim, that there have been trials of estrogen therapy following stroke, nonetheless, we agree with others1,2 that the mechanisms of action of estrogen to reduce at least cardiovascular events may only partly relate to its cholesterol-reducing effects and may also relate to other (eg, antioxidant) effects. In viewing older clinical trials of clofibrate, it is evident that one study3 with only 95 patients would not be powered to uncover a treatment effect on stroke. Another study4 showed no effect of the intervention on stroke. As has been pointed out by Davey Smith,5 as well as referenced in our article,6 there may be agent-specific effects that promote stroke as well as reduce stroke. The challenges implicit in mounting a trial specifically designed to look at the effects of statins in stroke reduction are formidable, not the least of which is the likelihood of reaching a cardiac end point prior to a stroke end point. Of interest, since the publication of our article, a second meta-analysis of clinical trials on the use of statins for reducing stroke has been published,7 and the authors agree with our conclusions.