OBJECTIVES: To assess the cardiac auscultatory skills of health care professionals before and after an actual (not simulated) patient-centered cardiac auscultation conference and to determine the effectiveness of this clinical teaching method on improving diagnostic proficiency. SUBJECTS AND METHODS: Seventy-eight participants at a conference on cardiac auscultation completed preconference and postconference examinations on their ability to diagnose a wide variety of heart sounds and murmurs. Among those tested, 46 (59%) were physicians: 17 (22%) were cardiologists, 19 (24%) were medical interns or residents, 7 (3%) were cardiology fellows, 2 (3%) were general practitioners, and 1 (1%) was a pediatrician. Thirty-two (41%) of this group of respondents were nonphysicians: 14 (18%) were nurse practitioners, 9 (12%) were nurses, 6 (8%) were physician assistants, 2 (3%) were medical students, and 1 (1%) was a nonmedical professional. All participants were tested on 14 clinically important cardiac auscultatory events directly recorded from actual (not simulated) patients and transmitted via wireless infrared stethophones. The auscultatory events tested were wide physiologic splitting of S2 in right bundle branch block, fixed S2 splitting and systolic murmur in atrial septal defect, midsystolic click and late systolic murmur of mitral valve prolapse, S4 gallop, S3 gallop, opening snap of mitral stenosis, innocent systolic murmur, systolic murmur of mitral regurgitation, systolic murmur of tricuspid regurgitation, systolic murmur of hypertrophic obstructive cardiomyopathy, continuous murmur of patent ductus arteriosus, pericardial friction rub, prosthetic valve sounds, and alternation of heart sounds, heart murmurs, and the S3 gallop in congestive heart failure. RESULTS: On the basis of the analysis of the 78 participants who completed preconferenee and postconference evaluations of the 14 selected cardiac auscultatory events, a preconference identification score of 26.3% and a postconference score of 44.7% were observed. This represents a statistically significant overall improvement in diagnostic proficiency (P<.001). CONCLUSIONS: Cardiac auscultatory skills among today's health care professionals are extremely poor, regardless of the level and/or type of training of the professional. An actual (not simulated) patient-centered teaching conference is an effective clinical method of improving cardiac auscultatory skills and diagnostic proficiency of health care professionals.