In any attempt to prevent the development of diseases in which known risk factors are involved, it is important to know the mode of correlation between risk factors and disease development. If the mode is an 'all-or-nothing' one, the maximum safe exposure limit is a vital parameter. If the correlation follows a continuous curve (as is the case in the lipid hypothesis of atherogenesis), a description of the natural distribution of risk factor compared with disease development is of interest. As such data are rare in unselected high-age groups and unknown in the very old without vascular disease, the distributions of total cholesterol and triglycerides, LDL- and HDL-cholesterol in 243 selected persons are communicated.