Literature on the diet of the Mexican-American population suggests that food consumption differs by nativity and duration of residence in the United States. However, no studies have used dietary pattern analysis to investigate these differences. To create dietary patterns, we used principal components analysis on food frequency questionnaire data for all adult respondents of the National Health and Nutrition Examination Survey years 2003-2006. Four dietary patterns were identified: Western, Healthy, Tomato/Tortilla, and Coffee/Sugar. Least squares means regression was used to test differences in adjusted mean dietary adherence scores among Mexico-born Mexican-American adults residing in the United States for <15 years (n=302), Mexico-born Mexican-American adults residing in the United States for ≥15 years (n=234), US-born Mexican-American adults (n=509), and US-born non-Hispanic whites (n=2,530) aged 18 to 69 years. Mean score for each diet pattern did not differ significantly by duration of US residence category in the Mexico-born Mexican-American population. However, in comparison to all Mexico-born Mexican Americans, US-born Mexican Americans had significantly lower score for the Tomato/Tortilla pattern, and significantly higher score for the Western pattern. Scores for the Healthy pattern were relatively low in all Mexican-American subgroups, indicating low adherence to the Healthy diet. However, Healthy diet scores of Mexico-born Mexican Americans among those aged ≤41 years appeared more similar to those of US-born non-Hispanic whites than US-born Mexican Americans. Education and policy action promoting healthy food access in Hispanic neighborhoods could help limit consumption of Western and Coffee/Sugar diet patterns and promote healthier choices in the Mexican-American population.