It has been suggested that gene-environment interaction is related to the risk of cancer. To evaluate departure from multiplicative effects between BRCA mutations and diet diversity in breast cancer (BC), a case-only study was carried out in a French-Canadian population including 738 patients with incident primary BC comprising 38 BRCA mutation carriers. Diet diversity was assessed using a validated food frequency questionnaire. Unconditional logistic regression analysis was performed to assess case-only odds ratio (COR) and 95% confidence interval (CI) while adjusting for age, body mass index, smoking, hormonal replacement therapy, and total energy intake. Ours results reveal a strong and significant interaction between BRCA mutations and vegetable and fruit diversity (COR = 0.27; 95%CI = 0.10-0.80; P = 0.03) when comparing the upper to the lower quartiles. The estimates for departure from multiplicative effects between BRCA mutations and total or other food groups' diversity were not supportive of the idea of a gene-environment interaction. The results of this study suggest that the combination of BRCA mutations and vegetable and fruit diversity may be associated with a reduced risk of BC.