Although associations of cholesterol and coronary heart disease (CHD) are well accepted, the association between cholesterol and stroke has been a subject of some confusion. Epidemiologic evidence suggests no association between plasma concentrations of cholesterol and stroke, and earlier clinical trials were also negative. Two early meta-analyses of clinical trials designed to evaluate the effects of cholesterol lowering on CHD concluded that cholesterol lowering had no effect. More recently newer, more potent and better tolerated agents (HMG-CoA reductase inhibitors, reductase inhibitors) have become available and have been tested for their efficacy in reducing cholesterol and CHD in both primary prevention and secondary prevention trials. Meta-analyses of these trials, in contrast to the earlier trials, reveal a powerful statistically significant effect to reduce stroke as well as CHD in secondary prevention (30%); the direction of the effect is the same in trials of primary prevention or trials that randomized patients with and without CHD (mixed primary and secondary prevention trials) where the risk reductions for stroke, although not reaching statistical significance are 11 and 30%, respectively. An important difference in the newer analysis is the availability of several trials of secondary prevention in which low density lipoprotein cholesterol was lowered 25-30% and in which CHD event reduction was similarly reduced by 30%. Mechanisms for stroke reduction likely involve retardation of plaque progression in the intracranial and extracranial carotid arteries, plaque stabilization, and, in addition, stroke may be reduced partly as a consequence of CHD reduction.