In this study we compared the diagnostic characteristics of the individual exercise electrocardiographic leads, 3 different lead sets comprising standard leads and the effect of the partition value in the detection of coronary artery disease (CAD). The diagnostic variable used was ST-segment depression at peak exercise, and the study population consisted of 101 patients with CAD and 100 patients with a low likelihood of the disease. The lead system used was the Mason-Likar modification of the standard 12-lead system and exercise tests were performed on a bicycle ergometer. The comparisons were performed by means of receiver-operating characteristic analysis and by determining sensitivities at a fixed 95% specificity. These properties, defined here as diagnostic capacity, were the most efficacious in leads I, -aVR, V4, V5, and V6. Diagnostic capacities in leads aVL, aVF, III, V1, and V2 were quite poor; statistical comparisons indicated significant differences between these leads and lead V5 (p < or = 0.0001 in each case). Use of the maximum value of ST-segment depression at peak exercise derived from all 12 leads produced a considerable decrease in the diagnostic capacity of the exercise electrocardiogram compared with lead V5. The exclusion of leads aVL, V1, and III improved the diagnostic capacity compared with the 12-lead set, but it was still smaller than that of lead V5. With use of a lead set with the 5 best leads increased the diagnostic capacity over other lead sets and over any individual lead. Further improvement was noted when a 50% smaller partition value was applied to leads I and -aVR than for the other leads (p = 0.041). In conclusion, this study suggests that use of leads I, -aVR, V4, V5, and V6 is the most influential when differentiating between patients with CAD and patients with a low likelihood of disease using peak exercise ST-segment depression. The effective use of leads I and -aVR requires the partition value applied for these leads to be 50% smaller than that used for the lateral precordial leads.