The prevalence and impact of atrial fibrillation (AF) versus sinus rhythm (SR) on outcomes in very severe aortic stenosis (vsAS) of the native valve is unknown. The aim of the study was to determine the prognostic significance of AF in vsAS. A total of 563 patients with vsAS (transaortic valve peak velocity ≥5 m/s) and left ventricular ejection fraction ≥50% were identified retrospectively. Patients were divided by rhythm at the time of index transthoracic echocardiogram (AF: n = 50 [9%] vs SR: n = 513 [91%]). Patients with AF were older (83.1 ± 7.5 vs 72.5 ± 12.2 y, p <0.001) and had no difference in gender distribution (p = 0.49) but had a higher Charlson co-morbidity index (2 [1,3] vs 1 [0,2], p = 0.01). There was no difference in transaortic peak velocity (5.3 ± 0.3 m/s vs 5.4 ± 0.4 m/s, p = 0.13) and left ventricular ejection fraction was comparable (63 ± 7 vs 66 ± 7%, p = 0.01). Age-, gender-, Charlson co-morbidity index-, and time-dependent aortic valve replacement (AVR)-adjusted overall mortality at 5 years was significantly higher in patients with AF than patients with SR (hazard ratio [HR] 1.88 [1.23 to 2.85], p = 0.003). AVR was associated with improved survival (HR = 0.30 [0.22 to 0.42], p <0.001), with no statistically significant interaction of AVR and rhythm (p = 0.36). Outcomes were also compared in the 2 SR:1 AF propensity-matched analyses (100 SR: 50 AF), with matching done according to age, gender, clinical co-morbidities, and year of echocardiogram. In the propensity-matched analysis, age-, gender-, and time-dependent AVR-adjusted all-cause mortality was higher in AF (HR 2.32 [1.41 to 3.82], p <0.001). In conclusion, AF was not uncommon in vsAS and identified a subset of patients at a much higher risk of mortality without AVR.