This study evaluated the reproducibility of four commonly utilized metabolic systems. The metabolic carts and associated electronic cycle ergometers were: (1) the Q‐PLEX I/Corival, (2) SensorMedics 2900/Ergometrics, (3) Medical Graphics Corporation (MGC) 2001/Mijnhardt, and (4) CPX/Medifit. Five individuals (mean age, 32 years; weight, 74 kg; height, 179 cm) experienced in graded exercise testing participated. Duplicate submaximal and maximal protocols were completed on separate days with all four systems. The submaximal power outputs averaged 85, 135, and 190 W and were each 4 minutes in duration. A maximal effort was achieved with a continuous ramp protocol (30 W/min) until exhaustion. All four systems demonstrated excellent reproducibility (usually 5% or less variation) for values reported during either submaximal or maximal exercise. These values were system dependent, however, as evident from a significant difference in submaximal and maximal pulmonary ventilation, oxygen consumption (VO2), carbon dioxide output, and respiratory exchange ratio between various systems. The average peak VO2varied by as much as 22% (45 to 55 ml‐kg‐min) among the systems. The group mean peak power output varied by as much as 39 W (296 to 335 W) among the cycle ergometers. Failure to measure a criterion pulmonary and metabolic functional capability will inevitably lead to the development of a divergent exercise prescription or training program in different locales using different equipment and will severely limit the interpretation and comparison of longitudinal or cross‐sectional studies.