Dipyridamole thallium scintigraphy has previously been shown to have prognostic value in the preoperative assessment of patients scheduled to undergo vascular surgery, but its effect on the long-term outcome is less well-defined. In the largest series to date, dipyridamole thallium scanning was performed in 360 patients before elective vascular surgery and survivors were followed for a mean of 31 months. In the 327 patients who underwent vascular surgery, operative death and nonfatal myocardial infarction occurred in 4.9 and 6.7%, respectively. A cardiac event (nonfatal myocardial infarction or cardiac death) occurred in 14.4% of patients with a transient thallium defect, as opposed to 1% with a normal scan (p < 0.001). Logistic regression analysis revealed that the best predictor of a perioperative event was the presence of a reversible thallium defect, elevating the risk by 4.3-fold. Late cardiac events occurred in 53 (15.2%) surgical survivors or nonsurgically treated patients. Patients with a fixed perfusion abnormality had a 24% late event rate, compared with 4.9% in those with a normal dipyridamole thallium study (p < 0.01). Cox analysis demonstrated that a fixed thallium defect was the strongest factor for predicting a late event and increased the relative risk by almost fivefold. A history of congestive heart failure was the only significant variable that contributed additional value to that of a fixed defect alone. Life-table analysis confirmed the strong relation of a fixed defect to cardiac event free survival (p < 0.0001).(ABSTRACT TRUNCATED AT 250 WORDS)