To determine correlates of survival in primary pulmonary hypertension, we compared 41 echocardiography-Doppler and nine catheterization parameters with outcome in 26 patients. Mean follow-up was 19.7 months in survivors; mean survival was 4.8 months in 16 nonsurvivors. Cox life-table univariate analysis correlated two echocardiographic, three Doppler, and three catheterization variables with poor survival (pc0.05), and xj analysis ensured the best critical values: severity of pericardial effusion, heart rate of more than 87 beats/min, pulmonic flow acceleration time of less than 62 msec, tricuspid early flow deceleration (T-DEC) equal to or less than -300 cm2/sec, mitral early flow-to-atrial flow velocity ratio (M-E/A) equal to or less than 1.0, catheterization cardiac index (CI) equal to or less than 2.3 1/min/m2, mean pulmonary artery pressure of more than 61 mm Hg, and diastolic pulmonary artery pressure of more than 43 mm Hg. Multivariate life-table analysis of noninvasive variables revealed the severity of pericardial effusion to be independently significant (p=0.006), whereas analysis of catheterization variables revealed cardiac index to be independently significant (p =0.014). Combined multivariate analysis did not differ from the noninvasive results alone. Categorical modeling of the eight significant variables split at their critical values (present or absent) revealed M-E/A, T-DEC, and CI to be independently significant by multivariate analysis (p=0.0014). Analysis of the five echocardiography-Doppler variables alone revealed M-E/A, T-DEC, and heart rate to be independently significant (p=0.0016). In both cases, mortality increased with the number of critical values reached. Thus, both echocardiography-Doppler and catheterization parameters yield valuable information regarding outcome in primary pulmonary hypertension. Noninvasive methods alone may be used to evaluate prognosis without sacrifice when the diagnosis of primary pulmonary hypertension has been made and subsequent invasive testing is unavailable, too expensive, or too hazardous to perform. (Circulation 1989;80:353-360 Ultrasound techniques now provide a noninvasive means of describing both the anatomic and hemodynamic derangements of the right heart and pulmonary circulation, including an accurate assessment of pulmonary pressures.'7-25 However, the usefulness of this information in predicting prognosis in PPH is unknown. Thus, we sought to evaluate the relations between two-dimensional and Dopplerechocardiographic variables and survival in PPH.