Dietary intake among other lifestyle factors influence blood pressure. We examined the associations of an "a priori" diet score with incident high normal blood pressure (HNBP; systolic blood pressure (SBP) 120-139 mmHg, or diastolic blood pressure (DBP) 80-89 mmHg and no antihypertensive medications) and hypertension (SBP ≥ 140 mmHg, DBP ≥ 90 mmHg, or taking antihypertensive medication). We used proportional hazards regression to evaluate this score in quintiles (Q) and each food group making up the score relative to incident HNBP or hypertension over nine years in the Atherosclerosis Risk of Communities (ARIC) study of 9913 African-American and Caucasian adults aged 45-64 years and free of HNBP or hypertension at baseline. Incidence of HNBP varied from 42.5% in white women to 44.1% in black women; and incident hypertension from 26.1% in white women to 40.8% in black women. Adjusting for demographics and CVD risk factors, the "a priori" food score was inversely associated with incident hypertension; but not HNBP. Compared to Q1, the relative hazards of hypertension for the food score Q2-Q5 were 0.97 (0.87-1.09), 0.91 (0.81-1.02), 0.91 (0.80-1.03), and 0.86 (0.75-0.98); p(trend) = 0.01. This inverse relation was largely attributable to greater intake of dairy products and nuts, and less meat. These findings support the 2010 Dietary Guidelines to consume more dairy products and nuts, but suggest a reduction in meat intake.