The purpose of this study was to determine whether the left ventricular response during exercise radionuclide angiography would be influenced by exercise protocol. One hundred twenty healthy volunteers (aged 18 to 40 years) performed upright bicycle exercise using 1 of 5 protocols. Ejection fraction was measured using first-pass radionuclide angiography. Exercise protocols were as follows: (1) graded exercise (25 W increase every 2 minutes) to fatigue, heart rate greater than 85% of age-predicted maximum, n = 53; (2) graded exercise to 85% of age-predicted maximal heart rate or to fatigue with heart rate less than 85% of age-predicted maximum, n = 26; (3) graded exercise to fatigue, with "exercise" imaging performed immediately after exercise, n = 15; (4) abrupt presentation of a supermaximal work load (400 W), n = 10; (5) graded exercise to a work load of 75 W preceding the abrupt presentation of a supermaximal work load (300 to 400 W), n = 16. Protocols 2 and 3, representing less than maximal stress, yield higher ejection fractions than Protocol 1 and may reduce the sensitivity of exercise radionuclide angiography. Protocols 4 and 5, representing supermaximal stress, yield lower ejection fractions than Protocol 1 and may reduce the specificity of exercise radionuclide angiography. Thus, exercise protocol has a significant influence on the left ventricular response during exercise radionuclide angiography.