From 1972 through 1975 a study of plasma lipid levels was conducted at the Columbia Medical Plan (a prepaid group practice in an upper-middle-class suburban community) in accordance with nationally standardized interview, blood-drawing and laboratory procedures of the Lipid Research Clinics Program. Data were obtained from a large group of volunteers from the plan as well as from subjects selected by random sampling from membership rolls. Of 2,591 fasting, nonpregnant adults (age greater than or equal to 20 years), 825 were volunteers and 1,766 were randomly selected subjects. Analysis of the plasma total cholesterol values indicated a possible association of volunteer status and higher educational levels with a lower plasma total cholesterol level. Age- and sex-specific comparisons confirmed this finding, although the magnitude of the differences was quite small from the standpoint of clinical risk. The data suggest that even within educational strata, self-selection for cholesterol screening was associated with a lower cholesterol level.