STUDY OBJECTIVE: To investigate whether neighbourhood characteristics are related to dietary patterns independently of individual level variables. DESIGN: A cross sectional analysis of the relation between neighbourhood median household income and food and nutrient intakes, before and after adjustment for individual level variables. SETTING: Four United States communities (Washington Co, MD; Suburban Minneapolis, MN; Forsyth Co, NC, and Jackson, MS). PARTICIPANTS: 13,095 adults aged 45 to 64 years participating in the baseline examination of the Atherosclerosis Risk in Communities (ARIC) Study, a prospective study of atherosclerosis. MEASUREMENTS AND MAIN RESULTS: Information on diet and individual level income was obtained from the baseline examination of the ARIC Study. Diet was assessed using a semi- quantitative food frequency questionnaire. Information on neighbourhood (census defined block groups) median household income was obtained from the 1990 US Census. Multilevel models were used to account for the multilevel structure of the data. Living in lower income neighbourhoods was generally associated with decreased energy adjusted intake of fruits, vegetables, fish, and increased intake of meat. Patterns generally persisted after adjustment for individual level income, but were often not statistically significant. Inconsistent associations were recorded for the intake of saturated fat, polyunsaturated fat, and cholesterol. Overall, individual level income was a more consistent predictor of diet than neighbourhood income. CONCLUSION: Despite limitations in the definition and characterisation of neighbourhoods, this study found consistent (albeit small) differences across neighbourhoods in food intake, suggesting that more in depth research into potential neighbourhood level determinants of diet is warranted.