A healthy diet and regular physical activity (PA) are delineated as healthy behaviours. Their implementation is associated with better health outcomes and improved quality of life. There is less evidence of a relationship between dietary patterns (DP) and PA, especially in adults aged ≥ 85. Hence, this cross-sectional study investigates the association between DP and PA in people of this age group, using the data from The Kawasaki Aging and Well-Being Project. Brief-type self-administered diet history questionnaire was used to estimate the intake of fifty-eight types of food. After energy adjustment, principal component analysis was performed to identify DP. PA was measured objectively using an accelerometer and subjectively using a questionnaire validated for this age group. Thousand participants (median age: 86·9 years, men: 49·9 %) were included in the analysis. Three major DP (DP1 ‘various foods’, DP2 ‘red meats and coffee’, DP3 ‘bread and processed meats’) were identified. DP1 ‘various foods’ was similar to DP previously named ‘healthy’ or ‘prudent’ and showed a positive association with PA time (PAT) as measured by accelerometer (B, 6·25; 95 % CI 0·13, 12·37) and relatively shorter sedentary behaviour (SB) time. DP2 ‘red meats and coffee’ and DP3 ‘bread and processed meats’ were negatively associated with PAT and positively associated with SB time. This study observed the relationship between diet and PA behaviours in adults aged ≥ 85, with healthier and more food-diverse DP associated with longer PAT and relatively unhealthy DP with shorter PAT.