PURPOSE--To study methodological aspects and results of cardiopulmonary exercise tests in elderly. METHODS--Twenty-five men (mean age 65 +/- 5 years) performed a cardiopulmonary exercise test using a bicycle ergometer and a progressive continuous work load increase protocol was employed. A computadorized system (2001 CAD/Net System-MGC), which includes a gas analyser and a pneumotacograph, was used for on line monitoring of oxygen and carbon dioxide expired fraction and also of pulmonary flow. RESULTS--We observed the following values of oxygen uptake, pulmonary ventilation and respiratory gas exchange: rest=4 +/- 1 ml/kg-1/min-1, 11 +/- 2 l/min-1 and 0.80 +/- 0,1, respectively: anaerobic threshold=12 +/- 3 ml/kg-1/min-1, 29 +/- 6 l/min-1 and 0.90 +/- 0.1, respectively: respiratory compensation point=18 +/- 4ml/kg-1/min-1, 47 +/- 1 l/min-1 and 1.07 +/- 0.1, respectively, and peak of exercise = 13 +/- 5 ml/kg-1/min-1, 76 +/- 18 l/min-1 and 1.21 +/- 0.2, respectively. The anaerobic threshold and the respiratory compensation point were achieved at 53 +/- 11 and 77 +/- 9//of peak oxygen uptake respectively. The elderly shown slightly increased values of the relation volume dead/tidal volume at rest with a slightly smaller decrease of its values during exercise in comparison to data obtained from young healthy untrained subjects. The heart rate prescription for exercise based on the heart rate reserve was higher than that based on the cardiorespiratory and metabolic responses (113-126bpm vs 96-114bpm). CONCLUSION--The cardiorespiratory and metabolic responses pattern in the elderly is quite heterogeneous. The heart rate prescription for exercise based on conventional stress tests seems to overestimate cardiorespiratory and metabolic capacity in healthy elderly men. The determination of anaerobic threshold and respiratory compensation point from cardiopulmonary exercise test data optimize exercise prescription for healthy elderly men