Objective To obtain reference values of aortic valve area (AVA) in a large population and to infer the risk of overestimating aortic stenosis (AS) when focusing on flow-corrected indices of severity. Methods We prospectively measured indices of AS in all consecutive echocardiograms performed in a large referral cardiac imaging laboratory for 1 year. We specifically analysed the distribution of AVA, indexed AVA and velocity ratio (Vratio) in patients with and without AS, the latter defined as the coexistence of valvular outflow obstruction (Vmax ≥2.5 m/s) and morphological findings of valve degeneration. Results 16 156 echocardiograms were analysed, 14 669 of which did not show valvular obstruction (peak jet velocity <2.5 m/s). In the latter group, AVA was 2.6±0.7 cm2 in 8190 studies with normal valves and 2.3±0.7 cm2 in 6479 studies with aortic sclerosis (AScl). There was a relatively wide overlap between values of AVA, indexed AVA and velocity ratio between studies of patients with AScl and AS. Values of AVA ≤1.0 cm2 were found in 0.5% of studies with normal valves and 1.8% of studies with AScl. These proportions were 3.1% and 9.3% for AVA ≤1.5 cm2, respectively. Vratio ≤0.25 were found in 0.1% of patients without obstruction. Risk factors for a small AVA in patients without obstruction were AScl, female sex, small body surface area, low ejection fraction and mitral regurgitation. Conclusions Normal values of continuity-equation derived AVA are smaller than previously considered. AVA values below cutoffs of moderate or severe AS can be found in patients without the disease. Flow-corrected indices may overestimate AS in patients with low gradients, particularly in the presence of well-identified risk factors.