BACKGROUND Exercise testing has long been employed in patients with hypertrophic cardiomyopathy (HCM), although concerns have constantly been expressed regarding its safety. This study reviews a large number of exercise tests performed in a community-based population with HCM, in terms of safety and clinical utility. METHODS We analyzed a total of 243 maximal symptom-limited cycloergometer exercise tests performed at our institution in 138 patients with HCM (age 42 +/- 14 years, M/F 99/39), who were followed systematically for 9.4 +/- 6.5 years. RESULTS In none of the 243 exercise tests did cardiac arrest, hemodynamic collapse or malignant arrhythmia occur, although 53 of the study patients (38%) had > or = 1 risk factors including previous cardiac arrest, recurrent syncope, malignant family history and resting left ventricular outflow obstruction. Early termination of the test was necessary in only 8 patients due to symptomatic hypotension with dizziness, but none had syncope. Mean predicted functional capacity achieved by the study group was 77 +/- 22%. Poor performance (< 60% of predicted functional capacity) was observed in 32 patients (23%), and it was associated with a NYHA functional class > 1 and an abnormal blood pressure response to exercise. Non-malignant arrhythmias occurred in 41 patients (30%), including multiple premature ventricular beats (PVB), paroxysmal atrial fibrillation, non-sustained ventricular (NSVT) and supraventricular tachycardia. The combined presence of multiple exercise-induced PVB and NSVT on Holter ECG had a 14% positive but a 97% negative predictive value for sudden death or cardiac arrest. CONCLUSIONS 1) Exercise testing is safe in a community-based population of patients with HCM, and provides useful information regarding functional capacity, efficacy of treatment, blood pressure response to exercise and inducible ischemia. Thus, ergometry should routinely be included in the standard evaluation and follow-up protocols of HCM patients. 2) Conversely, the utility of ergometry in the evaluation of the arrhythmic risk in HCM patients appears to be limited to the identification of low-risk patients.