Exercise trials in cardiology are often hindered by inconsistent approaches to exercise testing. These inconsistencies include the choice of exercise protocol, exercise end points, points of analysis, and absence or misuse of gas exchange data. Gas exchange techniques greatly enhance the accuracy with which cardiopulmonary function is assessed by exercise. Commonly used protocols are not always appropriate for all patients or all studies. Both cardiovascular disease and the exercise protocol can have an important impact on the relation between changes in work rate and oxygen uptake. Ramp protocols appear to offer the greatest promise for assessing cardiopulmonary function. Analyzing hemodynamic and gas exchange responses at several points submaximally, in addition to those at peak exercise, can add important information concerning the efficacy of a drug. A great deal of confusion continues to hinder the application of the gas exchange anaerobic threshold, and many of the commonly used testing end points are not reliable.