Background The national history of mild/moderate aortic stenosis (AS) has been previously reported in 3 studies. Aim We sought to assess the impact of the severity of AS in a group of patients initially managed medically. Methods A consecutive group of 399 patients with AS as assessed in our echolab (6.2.2002‐22.7.2003). Patients with a previous aortic prosthesis or operated within 3 months after diagnosis of AS were excluded. Patients were classified according to AS severity: * mild AS (AV area > 1.5‐2 cm2) * moderate AS (AV area 1‐1.5 cm2) * severe AS (AV area 0.8‐0.99 cm2) * very severe AS (AV area 0.8 cm2 had no impact on survival. The relative impact of the AS severity in addition to the clinical parameters (log Euroscore, age, NYHA, LVEJF) was assessed by the ROC model. The predictive performance was only slightly improved (by 2 ‐ 3%) by including the AS severity. Conclusion Survival in patients with very severe AS is reduced in comparison to patients with less severe AS. Survival in patients with less severe AS (AV area > 0.8 cm2) depends mainly on clinical characteristics, but not on degree of AS severity. There seems to be a place for randomized trials on the management of patients with less severe AS (AV area: 0.8‐1 cm2).