Introduction: Breast cancer is one of the main types of cancer in women. Previous studies assessed the relationship between diet and breast cancer risk, however, studies that evaluate the relationship between diet and mammographic findings are scarce. Aim: To evaluate the association between dietary patterns and mammographic findings of women monitored by a mastology service of the Brazilian Unified Health System. Materials and Methods: Cross-sectional study that evaluated data from 504 women. Two 24-hour dietary recalls were applied and dietary patterns were identified by principal component factor analysis. The mammography reports were classified into normal and altered mammographic findings. Logistic regression models were applied to evaluate the association between dietary patterns and mammographic findings. The analyses were performed in the software Stata and R and considered significant values of p <0.05. Results: Three dietary patterns was identified: “traditional Brazilian” (rice, beans, red meat, breads, oils and fats and coffees and teas), “Western” (sugar sweetened beverages, eggs, cakes, pies and cookies, fast-food snacks and, sweets and desserts) and “prudent” (whole grains, fruits, vegetables, tubers and their products, dairy and chicken). Women with less adherence (OR 0.54 CI 95% 0.30:0.95) to the “traditional Brazilian” dietary pattern had 46% lower odds for having altered mammographic findings. No associations were found between “Western” and “prudent” dietary patterns with mammographic findings. Discussion: The “traditional Brazilian" dietary pattern characterized by typical Brazilian foods such as rice and beans, was inversely associated with altered mammographic findings. This showed that women who have less adherence to this pattern had lower odds for having this outcome. Conclusion: We observed that women with less adherence to the “traditional Brazilian” pattern had lower odds for altered mammographic findings. We suggest that further studies should be performed, preferably with a longitudinal and case-control design.