Data linking low cholesterol with excess mortality have been gathered from both observational and clinical trials. The data are probably unrelated, as the levels of total cholesterol associated with excess mortality in observational studies are well below 160 mg/dl (4.1 mmol/l), whereas the levels achieved in clinical trials average around 230 mg/dl (5.9 mmol/l). Likely explanations of the association are that low cholesterol is a consequence of disease or is a confounder associated with other variables. The authors assess the relevance of the data to public health policy and medical practice. They conclude that evidence suggesting low cholesterol is a cause of excess mortality currently lacks breadth and rigor, and find no basis for changing cholesterol-management guidelines.