Abstract only Background: Evidence for the role of omega-3 fatty acids in the prevention of atrial fibrillation (AF) remain inconsistent, with some recent RCTs even suggesting possible harm. Whether long-term dietary intake of these fatty acids, as assessed using objective biomarkers, is related to AF is not known. Aims: To prospectively evaluate circulating and tissue levels of eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA), docosahexaenoic acid (DHA), and the sum of EPA and DHA (also known as the omega-3 index), with respect to incident AF. Methods: We used data from a global consortium of 13 prospective cohort studies with measurements of EPA, DPA, or DHA in adults (age≥18) identified through March 2021. Participating studies conducted de novo participant-level analyses using a prespecified analytical plan with harmonized definitions for exposures, outcomes, covariates, and subgroups. Associations were pooled using inverse-variance weighted meta-analysis. Results: Among 45,910 participants, a total of 6,229 incident cases of AF were ascertained, with median follow-up ranging from 0.9 to 29.1 years. In the multivariable analysis, per interquintile range (difference between the 90 th and 10 th percentiles for each fatty acid), DPA, DHA, and EPA+DHA ( Figure 1 ) were associated with 11%, 13%, and 9% lower incidence of AF, respectively ( P <0.01 for each). EPA levels were not associated with incident AF. Heterogeneity as measured using I 2 ranged from 0% for DPA to 56.7% for EPA+DHA. Associations were broadly consistent irrespective of baseline cardiovascular risk, global region, age, sex, or lipid fraction. Conclusion: Biomarkers of omega-3 fatty acids including DPA, DHA, and EPA+DHA demonstrated an inverse association with incident AF. In the absence of RCTs examining long-term dietary omega-3 intake and AF risk, our results do not suggest that higher levels of these fatty acids are associated with harm.