Background Dementia has no pharmacological cure. Therefore, lifestyle interventions targeting modifiable risk factors to reduce cognitive decline are of interest. This study examines the cross‐sectional relationships between two potentially protective behaviours: Mediterranean diet (MediDiet) adherence and physical activity (PA). Method Participants were recruited from the Sax Institute’s 45 and Up Study into the Maintain Your Brain trial. MediDiet adherence was assessed using the validated Mediterranean Diet and Culinary Index (MediCul) tool. The 50‐item tool consists of 17 sub‐categories focusing on key aspects of the MediDiet. Leisure time PA was assessed by a standard questionnaire and intensity was quantified using the BORG Rating of Perceived Exertion (RPE) scale, modified for strength and aerobic activities. Associations between the MediDiet and PA were investigated using hierarchical linear regression and analysis of covariance. Result 6236 participants 55‐77 years; mean (SD)=65.0 (5.8) completed baseline assessments and were included. Mean (SD) MediCul score was 53.2 (13.0)/100), indicating low adherence to the MediDiet. Only 5% of participants achieved a score consistent with better cognitive outcomes in The PREDIMED study. Almost one‐half of participants (48.4%) met aerobic PA (150 min/week) but less than one‐quarter (24.2%) met resistance training (RT) recommendations (2 days/week). Unadjusted MediCul score explained a small but significant amount of the variance for light (1.0%) and moderate‐vigorous (MV) (3.1%) PA, both p<0.001. For light PA, the final model, including MediCul, age, sex, BMI, CAGE (alcohol use) score and diabetes explained 2.8% of the variance. For MV PA, the final model including MediCul, age, sex, BMI, CAGE, depression, diabetes and education explained 10.9% of the variance. A 10‐point higher MediCul score was associated with an additional 3.3 seconds of light PA/wk and additional 7.5 seconds of MV aerobic PA/wk (both p<0.001). Additionally, MediCul score was significantly higher in participants engaging in 2+days/wk of RT compared to 1 or fewer days/wk (56.6/100 vs. 52.2/100, respectively; p<0.001). Conclusion Both aerobic and RT PA are significantly but weakly associated with better diet, but the clinical meaningfulness, as well as any causal nature, of these relationships requires further exploration. The outcomes of the MYB trial will contribute substantively to this question.