OBJECTIVES: To examine the patterns and influences on repeated emergency department attendance among frail older people with deteriorating health. STUDY DESIGN: Multicentre prospective cohort study (International Access Rights and Empowerment II study) with convergent mixed methods design. METHODS: Eligible patients were aged ≥65 years, with Clinical Frailty Score ≥5, and ≥1 hospital admission or ≥2 acute attendances in the previous 6 months. Questionnaires were administered to participants over 6 months and we extracted clinical data from the medical records. We conducted modified Poisson multivariable regression analysis to identify factors associated with repeated emergency department attendance (≥2 over 6 months) and thematic analysis of qualitative interviews. RESULTS: A total of 90 participants were recruited. The mean age was 84 years, and 63% were women. Of 87 participants, 21 experienced repeated emergency department attendance. Severe and/or overwhelming pain (adjusted prevalence ratio 2.44, 95% confidence interval 1.17-5.11), greater number of comorbidities (1.32, 1.08-1.62), ≥10 community nursing contacts (2.93, 1.31-6.56), and a total of ≥2 weeks spent in hospital during the previous 6 months (2.91, 1.24-6.84) were associated with repeated attendance. From 45 interviews, we identified influences on emergency department attendance: 1. inaccessibility of community healthcare; 2. perceived barriers to community healthcare seeking; 3. perceived benefits of hospital admission; 4. barriers to recovery during previous hospital admission (unsuitable food, inactivity); and 5. poorly coordinated transitions between settings. CONCLUSIONS: We identified missed opportunities to optimise older people's recovery during hospital admission, such as improved food and a timely and coordinated discharge, which may reduce reattendances. Proactive care in the community with systematic assessment of symptoms may be required, particularly for those with multimorbidity.