To evaluate the reproducibility of ejection fraction (EF) and regional wall motion (RWM) analyses by rest and exercise equilibrium radionuclide ventriculography (RNV) in the presence of coronary artery disease (CAD), 18 patients underwent two maximum, multistage supine bicycle exercise studies separated by an interval of 2 weeks. There were no significant differences in EF between the two studies, both at rest (56.0 +/- 13.8% vs 58.2 +/- 11.7%, p = NS) and with exercise (51.1 +/- 17.6% vs 54.3 +/- 17.6%, p = NS) and a highly significant correlation was shown between the two groups of values (rest r = 0.90, exercise r = 0.93, p less than 0.001). There was no significant difference in the change from rest to exercise (-4.9 +/- 12.0% vs -3.8 +/- 11.5%, p = NS) between the two studies and the correlation was highly significant (r = 0.69, p less than 0.01). The interstudy variabilities were 2.2 +/- 6.1% and 1.2 +/- 7.3% for rest and exercise, respectively, and 2.0 +/- 9.2% for the change from rest to exercise. Ninety-four percent of both rest and exercise regions had similar RWM. Eighty-one percent of the abnormally contracting regions were common to both exercise studies. Utilizing conventional criteria for the diagnosis of CAD, 11 patients had abnormal EF response and nine had abnormal RWM response to exercise on both studies. Combining EF and RWM criteria resulted in the diagnosis of CAD in 15 patients in both studies. We conclude that: (1) there were no significant differences in rest and exercise radionuclide EF and RWM between two supine bicycle exercise studies performed 2 weeks apart in patients with stable CAD and there were significant correlations between the two studies; (2) despite these correlations, the interstudy variabilities emphasize the need for the inclusion of reproducibility studies in all evaluations of interventions by exercise radionuclide ventriculography; and (3) the variations in EF and RWM response to exercise result in lack of uniformity between the two studies regarding the diagnosis of CAD based on conventional RNV criteria.