This study compares hemodynamic, metabolic, and gas exchange responses, catecholamine levels, and symptoms in 35 male patients with chronic heart failure (CHF) ([mean ± SD] age 53 ± 11 yr; ejection fraction 24± 11%) during three differently graded exercise test protocols. On three consecutive days patients performed cycle ergometry supine, with prolonged steps (prol BE) and right heart catheterization, ramplike cycle ergometry sitting (ramp BE), and ramplike treadmill walking (TMW). As in routine clinical practice, the prol BE was terminated when pathologic central hemodynamics and/or increased symptomology occurred, and ramp BE and TMW due to increased symptomology and/or physician's decision. During prol BE at ventilatory threshold (VT) the ˙VO2 (8.6 ± 1.8 ml·kg-1·min-1) was lower than during ramp BE (9.3± 2.1 ml·kg-1·min-1) (P < 0.017) and TMW (11.8 ± 2.3 ml·kg-1·min-1)(P < 0.0001). Prol BE, ramp BE, and TMW also differed significantly with respect to ventilation (22 ± 71·min-1; 26 ± 6 l·min-1; 29 ± 7 l·min-1;P < 0.01) and heart rate (100 ± 15 beats·min-1; 103 ± 18 beats·min-1; 110± 16 beats·min-1; P < 0.017). No differences were found in lactate levels, catecholamine levels, and ratings of leg fatigue and dyspnea. At test termination, the peak ˙VO2 during prol BE(10.8 ± 3.3 ml·kg-1·min-1) was lower than during ramp BE (13.3 ± 4.1 ml·kg-1·min-1)(P < 0.0001) and TMW (14.7 ± 3.4 ml·kg-1·min-1) (P < 0.0001). Peak norepinephrine value during ramp BE (4.531 ± 2.788 nmol·l-1) was higher than during prol BE (3.707 ± 2.262 nmol·l-1) (P < 0.001). Among the three tests, no significant differences were found for peak values of heart rate, lactate, and ratings of dyspnea. Although the ˙VO2·kg-1 at VT was significantly higher during ramp BE and TMW compared to prol BE (P< 0.001), the values expressed as a percent of peak˙VO2·kg-1 were significantly lower (70 ± 4%; 72 ± 6%; 79 ± 3%; P < 0.017). A systematic effect on aerobic capacity with reduced peak values during ramp BE and TMW was demonstrated when test termination was based primarily on pathological findings of central hemodynamics from prol BE.