In previous studies to predict future cardiac death of patients undergoing evaluation for renal transplantation, noninvasive or invasive testing of all, or nearly all, patients has been used. In an attempt to decrease the cost of cardiac risk assessment, we prospectively used a two-tiered cardiac risk assessment algorithm on 189 consecutive patients referred for renal transplant evaluation. First, patients were stratified by clinical characteristics of age > or = 50 years, history of angina, insulin-dependent diabetes, congestive heart failure, or abnormal electrocardiogram (excluding left ventricular hypertrophy). Patients having none of these risk factors (n = 94) were considered at low risk for cardiac events and underwent no further cardiac evaluation. Patients with one or more of the cardiac risk factors (n = 95) were considered to be in a high-risk group and were required to undergo further evaluation with thallium myocardial scintigraphy. Follow-up of patients was for 46 +/- 16 months. Cardiac mortality was significantly higher in the clinical high-risk group compared with the clinical low-risk group (17% v 1%, respectively; P < 0.001). Further cardiac risk stratification was evident by thallium myocardial scintigraphy. Patients with reversible thallium defects had significantly higher cardiac mortality rates than patients with no thallium defects (23% v 5%, respectively; P < 0.05). Fixed thallium defects also had predictive value for cardiac mortality (29%,; P < 0.05), but deaths in this fixed defect group tended to occur later in the follow-up period. The initial clinical stratification obviated the need for further noninvasive or invasive testing in nearly half of the renal transplant candidates.(ABSTRACT TRUNCATED AT 250 WORDS)