In this article, we discuss a problem in the most recent American Heart Association guideline on secondary stroke prevention that apparently arose from the rules of evidence imposed on the guideline panel. We are told by the cochair of the panel that American Heart Association rules about guidelines for secondary prevention prohibited consideration of primary prevention studies and secondary analyses of secondary prevention studies. However, evidence-based medicine should consider all the best external evidence available and also clinical judgement. The most important problem in the guideline was the recommendation that B vitamins to lower homocysteine do not prevent recurrent stroke. When considering all the best external evidence, it is clear that B vitamins do prevent stroke, but in the early secondary stroke prevention studies, the benefit of B vitamins in participants with good renal function was apparently offset by harm from cyanocobalamin among participants with renal failure (level B-R evidence). We review the evidence that B vitamins should be used to prevent stroke, both in primary and secondary stroke prevention (class 2a recommendation). We also review issues in folate metabolism that require further study, with regard to the form of folate to be used for stroke prevention. We recommend that the guideline be revised to say that B vitamins to lower homocysteine prevent stroke and that methylcobalamin or hydroxycobalamin should be used instead of cyanocobalamin.