BACKGROUND: The measurements used to define pulmonary hypertension (PH) etiology, pulmonary arterial wedge pressure (PAWP), and left ventricular end-diastolic pressure (LVEDP) vary in clinical practice. We aimed to identify clinical features associated with measurement discrepancy between PAWP and LVEDP in patients with PH. METHODS: We extracted clinical data and invasive hemodynamics from consecutive patients undergoing concurrent right and left heart catheterization at Vanderbilt University between 1998 and 2014. The primary outcome was discordance between PAWP and LVEDP in patients with PH in a logistic regression model. RESULTS: We identified 2,270 study subjects (median age, 63 years; 53% men). The mean difference between PAWP and LVEDP was -1.6 mm Hg (interquartile range, -15 to 12 mm Hg). The two measurements were moderately correlated by linear regression (R = 0.6, P < .001). Results were similar when restricted to patients with PH. Among patients with PH (n = 1,331), older age (OR, 1.77; 95% CI, 1.23-2.45) was associated with PAWP underestimation in multivariate models, whereas atrial fibrillation (OR, 1.75; 95% CI, 1.08-2.84), a history of rheumatic valve disease (OR, 2.2; 95% CI, 1.36-3.52), and larger left atrial diameter (OR, 1.70; 95% CI, 1.24-2.32) were associated with PAWP overestimation of LVEDP. Results were similar in sensitivity analyses. CONCLUSIONS: Clinically meaningful disagreement between PAWP and LVEDP is common. Atrial fibrillation, rheumatic valve disease, and larger left atrial diameter are associated with misclassification of PH etiology when relying on PAWP alone. These findings are important because of the fundamental differences in the treatment of precapillary and postcapillary PH.