OBJECTIVE: Consumption of fish and marine n-3 polyunsaturated fatty acids (PUFA) may be associated with a lower risk of atrial fibrillation (AF), but results have been inconsistent. The aim was to investigate this further by measurements of marine n-3 PUFA in adipose tissue. DESIGN: Cohort study. SETTING: A total of 57 053 Danish participants 50-64 years of age were enrolled into the Diet, Cancer and Health Cohort Study. PATIENTS: A randomly drawn subcohort of 3440 participants with available data from baseline adipose tissue biopsies. INTERVENTIONS: Exposure was the adipose tissue content of marine n-3 PUFA, which reflects the endogenous exposure and is also an objective marker of the long-term dietary intake. MAIN OUTCOME MEASURES: Incident AF during follow-up. RESULTS: 179 cases of AF occurred over 13.6 years. Multivariate, sex-stratified Cox proportional hazards regression analyses using cubic splines showed a monotonic, negative, dose-response trend, but not statistically significant association, between total marine n-3 PUFA in adipose tissue and incident AF. A similar trend towards a lower risk of AF was seen in the second (HR 0.87, 95% CI 0.60 to 1.24) and third tertiles (HR 0.77, 95% CI 0.53 to 1.10) of marine n-3 PUFA compared with the lowest tertile. Similar trends, but also not statistically significant, were found separately for eicosapentaenoic, docosahexaenoic and docosapentaenoic acids. CONCLUSIONS: There was no statistically significant association between the content of marine n-3 PUFA in adipose tissue and the development of AF; however, data showed a monotonic, negative dose-response trend suggestive of a negative association.