We examined the relationship between intake of food group (and supplement) sources of folate and plasma folate and homocysteine concentrations among 885 elderly subjects in the Framingham Heart Study. Dietary data were collected by food-frequency questionnaire, and blood samples analyzed for folate and homocysteine concentrations. Top contributors to total folate intake were ranked. Mean folate intake, plasma folate and homocysteine concentrations were estimated for users vs. non-users of supplements, and key foods--those which both contribute to total folate intake and are known to be good sources of folate--and examined statistically with adjustment for age, gender and total energy intake. Plasma folate and homocysteine concentrations were also determined by quintile of intake frequency for breakfast cereals and for fruits and vegetables. Plasma folate was significantly greater and homocysteine lower in women than in men. Despite somewhat greater plasma folate concentrations with age, homocysteine was significantly higher in those over 80 y of age than in younger subjects. Major contributors to folate intake were cold breakfast cereals (13.3%), multivitamins (12.8%) and orange juice (12.4%). Users of supplements, breakfast cereals, or green leafy vegetables had significantly greater plasma folate and lower homocysteine levels than non-users. Plasma folate concentration was also greater in those who drank orange juice. We identified clear dose-response relationships for both plasma folate and homocysteine with increased quintile of breakfast cereal and of fruit and vegetable use. Frequent consumption of these foods is associated with higher folate and lower homocysteine concentrations.