Patients with chronic stable angina pectoris may present with either fixed or variable threshold symptoms. To evaluate the diagnostic value of ambulatory Holter monitoring for the detection of coronary artery disease (CAD) in patients with variable threshold angina, 216 consecutive candidates for coronary angiography were investigated prospectively. For comparison, a group of 55 consecutive patients with fixed threshold angina was studied under the same conditions. Patients with prior myocardial infarction or angiographically documented CAD were excluded. Within 4 months of Holter monitoring, the advised coronary angiography was performed in 77% of the patients with variable threshold angina and in 89% of the patients with fixed threshold angina (p less than 0.05). The prevalence of CAD was markedly lower in patients with variable threshold angina compared to patients with fixed threshold angina (54 vs 90%, p less than 0.001). CAD patients of both subgroups, however, did not differ significantly with respect to the number of obstructed vessels, the Gensini coronary score, the number with impaired left ventricular function (ejection fraction less than 50%) or the duration of ischemic episodes during Holter monitoring. Diagnostic accuracy of Holter monitoring did not differ between variable and fixed threshold angina groups (67 vs 78%). In 91% of the patients results obtained by Holter monitoring could be compared to the results of a bicycle stress test. In patients with fixed threshold angina the diagnostic accuracy was similar for both tests (80 vs 80%). In patients with variable threshold angina, the diagnostic accuracy of Holter monitoring exceeded that of the exercise stress test (68 vs 55%, p less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)