Abstract Background Elite athletes frequently demonstrate outspoken bi-atrial dilation, even exceeding volumes seen in patients with atrial fibrillation (AF). Hence, distinguishing atrial exercise-induced adaptation from maladaptation or atrial pathology cannot be done using volumes only. However, data on bi-atrial strain in athletes compared with AF patients investigated using cardiac magnetic resonance imaging (CMR). Methods We performed a retrospective cross-sectional analysis of CMR data comparing three groups: (1) elite athletes included in the ELITE-cohort at the Amsterdam UMC (> 16 years of age and train ≥ 10 hours/week at (inter)national or Olympic level, (2) paroxysmal- (PAF) and persistent AF patients awaiting PVI and (3) healthy controls. All participants underwent CMR in sinus rhythm. Left- and right atrial (LA/RA) reservoir (RES), conduit (CD) and contractile (CT) strain were analysed by automatic feature tracking (FT). Results We included a total of 340 elite athletes (median age 28 years [17, 67]; female n=145 (42.6%)), 103 AF patients (median age 60 [29, 75], female n=36 (35%), PAF n=74 (71.8%), persistent AF n=29 (28.2%), and 85 healthy controls (median age 28 [16, 72], female n=39 (45.9%). Elite athletes had similar indexed maximal left atrial volumes (LAVI) compared to AF patients (49.6 ml/m2 [41.9, 58.3] vs 47.9 ml/m2 [40.5 ml/m2], p = .25), but larger indexed maximal right atrial volumes (RAVI) (59.3 ml/m2 [50.7, 70.7] vs 44.7 ml/m2 [37.0, 55.7], p < .001). Elite athletes had lower LA RES strain (42.8% [36.7. 49.2]) than controls (47.5% [40.6, 54.3], p < .001) but a higher LA RES than AF patients (34.7% [27.4, 41.6], p < .001). Athletes demonstrated higher LA CD/CT ratio (2.2 [1.8, 2.8]) than healthy controls (1.8 [1.4, 2.2], p < .001) and AF patients (1.4 [1.1, 1.9], p < .001) caused by a higher LA CD in athletes (29.2% [24.7, 34.2] compared to healthy controls (29.1% [24.0, 36.2], p < .001) and AF patients (20.1% [16.2, 25.6], p < .001) and a lower LA CT in both athletes (13.0% [22.1, 15.8] and controls (20.1% [16.2, 25.6], compared to healthy controls (17.2% [14.1, 19.9], p < .001). Elite athletes had lower RA RES strain (51.5% [42.8. 63.4]) than controls (57.3% [48.0, 73.5], p = .001) and higher RA RES than AF patients (44.7% [34.1, 55.0], p < .001). AF patients had lower RA CD (28.3% [20.1, 36.8] than elite athletes (35.9% [28.9, 45.5], p < .001) and healthy controls (41.9% [32.4, 51.1], p < .001). Elite athletes had higher RA CD/CT ratio than AF patients (2.5 [1.9, 3.2] vs 1.9 [1.2, 2.4], p < .001). Conclusion Compared with healthy controls and patients with AF, athletes demonstrate unique bi-atrial characteristics. While LAVI is similar in both elite athletes and AF patients, elite athletes demonstrate larger RAVI and a unique strain profile characterized by a higher bi-atrial CD/CT ratio compared with AF patients, but a reduced bi-atrial RES strain relative to healthy controls yet elevated compared to AF patients.Differences in bi-atrial strain