To determine the relation between myocardial ischemic indexes on exercise testing and on ambulatory Holter recording, 60 patients with stable coronary artery disease who exhibited an ischemic response to both testing procedures were studied. All patients performed a Bruce protocol exercise test and underwent 24-hour Holter recording within 2 weeks without antianginal medications. Mean exercise duration was 7.4 +/- 2.8 minutes, mean heart rate at 1-mm ST depression was 118 +/- 20 beats/min and mean maximal ST depression during exercise was 2.2 +/- 1 mm. During Holter recording the average number of ischemic episodes was 4.7 +/- 2.6 per patient, mean duration of daily ischemia was 62 +/- 54 minutes, mean maximal ST depression was 3.2 +/- 1.3 mm and average heart rate at 1-mm ST depression was 93 +/- 17 beats/min. Overall, the correlations between the ischemic indexes on both testing procedures were very weak (mean r2 = 0.054). The only exercise variable that had a significant correlation (p < 0.05) with all Holter variables was heart rate at 1-mm ST depression, yet it correlated very weakly (0.064 less-than-or-equal-to r2 less-than-or-equal-to 0.125) with most Holter covariates and had a better correlation (r2 = 0.256) only with average heart rate at 1-mm ST depression during Holter. Thus, ischemic indexes on exercise testing cannot accurately predict ischemic indexes on ambulatory Holter recording in patients with stable coronary artery disease who exhibit ischemic changes on both tests. These findings indicate that a great portion of the variability of Holter ischemic indexes is dependent on factors other than those represented by exercise testing indexes, and that Holter recording in the ambulatory setting might provide clinically relevant information in addition to that obtained by exercise testing in selected subsets of patients.