BACKGROUND Poor performance by physicians-in-training and interobserver variability between physicians has diminished clinicians' confidence in the value of the fourth heart sound (S4). OBJECTIVE We sought to determine if accurate auscultation of an S4 improves with advancing levels of experience. METHODS We performed a prospective study of 100 patients undergoing left heart catheterization. Patients underwent blinded auscultation by 4 physicians (each from 1 of 4 different levels of experience), computerized acoustic cardiography, measurement of B-type natriuretic peptide (BNP) levels, echocardiography for measurement of left ventricular ejection fraction (LVEF), and cardiac catheterization for measurement of left ventricular end-diastolic pressure (LVEDP). RESULTS While cardiology fellows', residents', and interns' auscultatory findings demonstrated no significant agreement with acoustic cardiography, an S4 auscultated by cardiology attendings had moderate diagnostic accuracy with acoustic cardiography (odds ratio [OR]: 2.31; receiver-operating-characteristic [ROC] area: 0.60). The sensitivities of the S4 were low (39%-46%) for identifying patients with abnormal measures of left ventricular filling pressure (BNP and LVEDP, respectively), and the specificities were fair (76%-80%) with acoustic cardiography. The S4 was not associated with abnormal LVEF. None of the ausculatory groups performed as well as acoustic cardiography in separating patients based on objective measures of left ventricular filling pressure. Acoustic cardiography had the lowest (superior) negative likelihood ratios compared to any ausculatory group. CONCLUSIONS The S4 auscultated by cardiology attendings demonstrated superior diagnostic test characteristics compared with internal medicine housestaff and cardiology fellows. Correlations between the S4 and measures of ventricular filling pressure were superior for acoustic cardiography compared to the auscultator groups.