Contemporary outcomes for aortic stenosis (AS) and the association between physician-assessed AS severity and quantitative parameters is poorly understood. We aimed to evaluate AS natural history, compare outcomes for physicians’ AS assessment vs. quantitative parameters, and identify AS parameters with the most explanatory power. We ascertained physician-assessed AS severity, echocardiographic parameters, and clinical data for 546,769 patients from 2008-2018, examined multivariable associations of physician-assessed AS severity and number of quantitative severe AS parameters with death, cardiovascular hospitalization, and aortic valve replacement, and estimated the relative contribution of different quantitative AS parameters on outcomes. Among 49,604 AS patients (mean [SD] age 77  years), 17.6% had moderate, 3.6% moderate-severe, and 9.4% severe AS. During median 3.7 [IQR 1.7-6.8] years, physician-assessed AS severity strongly correlated with outcomes, with moderate AS patients tracking closest to mild AS, and moderate-to-severe AS patients more comparable to severe AS. Although the number of quantitative severe AS parameters strongly predicted outcomes (adjusted HR [95% CI] for death 1.40 [1.34-1.46], 1.70 [1.56-1.85], and 1.78 [1.63-1.94] for 1, 2, and 3 parameters, respectively), aortic valve area <1.0cm2 was the most frequent severe AS parameter, explained the largest relative contribution (67%), and was common in patients classified as moderate (21%) or moderate-severe (56%) AS. Physician-assessed AS severity predicts outcomes, with cumulative effects for each severe AS parameter. Moderate AS includes a wide spectrum of patients, with discordant AVA <1.0cm2 being both common and predictive. Better identification of non-classical severe AS phenotypes may improve outcomes.