The utility of metabolic gas exchange measurements in evaluating the severity and determinants of exercise limitation was studied during upright symptom-limited cardiopulmonary exercise in 135 consecutive patients with hypertrophic cardiomyopathy (HC) and 50 healthy age- and gender-matched volunteers. Peak oxygen consumption (VO(2)) was less than predicted (age, gender, and size) in 99% patients. Peak VO(2) was significantly associated with New York Heart Association functional class; however, there was considerable overlap of peak VO(2) between classes I and III (70 +/- 15%, 56 +/- 15%, 35 +/- 11%, respectively). Patients with abnormal blood pressure responses and patients with chronotropic incompetence during exercise had lower percent-predicted peak VO(2) than patients with normal blood pressure and heart rate responses during exercise (p = 0.0001 and p <0.001, respectively). Percent-predicted peak VO(2) was similar in patients with and without resting left ventricular outflow obstruction. Of those patients with resting gradients, however, there was a strong inverse correlation between the magnitude of the gradient and peak VO(2) (r = 0.5; p <0.001). In conclusion, peak VO(2) is significantly related to New York Heart Association functional class in this group of patients with HC, but peak VO(2) is a superior measure of cardiovascular performance in individual patients. Our peak VO(2) data indicate that mechanical obstruction has an adverse pathophysiologic effect on functional capacity and provide the rationale to support treatments aimed at gradient reduction. Low peak VO(2) characteristics including those with normal or near-normal left ventricular wall thickness suggests that measurement of peak VO(2) may aid in the differential diagnosis between HC and athlete's heart.