In addressing the potential for any hazards associated with lowering serum cholesterol, a recent meta-analysis by Law and colleagues (Br Med J 1994;308:373) concluded that that there was no evidence that a low or reduced serum cholesterol concentration increases mortality from any cause other than haemorrhagic stroke. In this review, the evidence for this unexpected association will be addressed, and possible mechanisms discussed. Overall implications of the phenomenon, particularly for the hypertensive patient, will be addressed in the context of the anticipated consequences of lipid-lowering therapy for all stroke, haemorrhagic and non-haemorrhagic, as well as ischaemic heart disease (IHD).