Background: Global longitudinal strain (GLS) is a sensitive marker of myocardial dysfunction and atrial reservoir function. We sought to evaluate its value for predicting atrial fibrillation (AF) in the general population. Methods: Participants from the Copenhagen City Heart Study examined with echocardiography, including speckle tracking analyses, were included. The endpoint was AF obtained through national registries. Proportional hazards Cox regression was applied, including multivariable adjustments made for CHADS₂ and CHARGE-AF risk factors. Abnormal GLS was defined as >—18%. Results: The data from 1,309 participants were analyzed. Of those, 153 (12%) developed AF during a median follow-up time of 15.9 years. The follow-up was 100%. The mean age was 57 years, 38% had hypertension, and GLS was − 18%. In unadjusted analysis, GLS was a univariable predictor of outcome (1.08 (1.04–1.13), p < 0.001, per 1% absolute decrease), but did not remain an independent predictor after adjusting for neither CHADS₂ nor CHARGE-AF risk factors. However, hypertension modified the relationship between GLS and AF (p for interaction = 0.010), such that GLS only predicted AF in subjects without hypertension. In participants without hypertension, GLS remained an independent predictor of AF after adjusting for CHADS₂ and CHARGE-AF (HR = 1.11 (1.03–1.20) and HR = 1.09 (1.01–1.19), respectively). In these participants, an abnormal GLS was associated with a more than twofold increased risk of AF (HR = 2.16 (1.26–3.72). The incidence rate was 3.17 and 6.81 per 1000 person-years for normal vs. abnormal GLS, respectively. Conclusion: Global longitudinal strain predicts AF in individuals without hypertension from the general population, independently of common risk scores. Graphical abstract: [Figure not available: see fulltext.]