Previous reports have indicated that coffee consumption may increase serum cholesterol levels. We studied the effects of coffee prepared by two common brewing methods (filtering and boiling) on serum lipid levels in a 12-week randomized trial involving 107 young adult subjects with normal serum cholesterol levels. After a three-week run-in period during which they all consumed filtered coffee, the participants were randomly assigned to one of three groups receiving four to six cups of boiled coffee a day, four to six cups of filtered coffee a day, or no coffee, for a period of nine weeks. As compared with the change from base line in the filtered-coffee group, the serum total cholesterol level increased after the consumption of boiled coffee by 0.48 mmol per liter (95 percent confidence limits, 0.13 and 0.83), and the low-density lipoprotein cholesterol level increased by 0.39 mmol per liter (95 percent confidence limits, -0.04 and 0.82). There was no significant difference in the change in serum total or low-density lipoprotein cholesterol levels between the filtered-coffee group and the group that drank no coffee. The levels of high-density lipoprotein cholesterol and apolipoproteins were not affected by boiled or filtered coffee. We conclude that drinking filtered coffee does not affect serum lipid levels. The consumption of boiled coffee, however, has an effect on serum cholesterol levels amounting to a mean net increase of 10 percent of the base-line level after nine weeks.