The relationship between coffee consumption and serum cholesterol was investigated in a cohort of 5,858 Japanese males born in 1900-1919 and living in Hawaii in 1965 who are currently followed by the Honolulu Heart Program. Data on coffee consumption, other dietary variables from a 24-hour dietary recall, and other potentially confounding variables collected in 1965 were correlated with serum cholesterol at that examination and at examination six years later. The mean coffee and tea consumption was 3.4 and 1.8 cups/day, respectively. Those consuming no coffee had a mean serum cholesterol of 210 mg/dl, while that of those drinking 9+ cups/day was 220 mg/dl (no such relationship was apparent with tea or cola). The relationship of coffee consumption and serum cholesterol with potentially confounding variables including body mass index, cigarette smoking, diastolic blood pressure, alcohol consumption, physical activity index, serum glucose, serum uric acid, education, age, and fat consumption was examined. When these variables were entered into a multiple regression equation with coffee consumption, a significant relationship between coffee consumption and serum cholesterol (p less than 0.001) persisted, as did that between baseline coffee consumption and serum cholesterol six years later (p less than 0.001). There was no significant relationship between tea or cola, the other major caffeine contributors to the diet, and baseline serum cholesterol. Thus, this analysis indicates a significant positive relationship between coffee consumption and serum cholesterol which is not present with other sources of caffeine.