One hundred and twenty-lead body surface potential maps (BSPMs) were recorded at rest, at immediate cessation of exercise and after 1 (early) and 5 minutes (late) of recovery in 14 patients with isolated, critical, left anterior descending (LAD) coronary artery stenosis. Exercise endpoints, at an average peak rate of 98 +/- 13, were usual pain worsening in 13 LAD patients, and diagnostic ST depression in lead V5 in 1 patient. Twelve patients also had positive thallium scans. BSPMs were also recorded in 8 normal subjects who exercised to peak heart rates similar to those of the LAD subjects. Spatially, there were similar exercise changes in QRS and ST-segment integral patterns over the precordium and inferior torso in both groups. These were transient in the control group but persisted to late recovery in the LAD group, particularly for ST integral. Quantitatively, multivariate analysis revealed significant temporal differences between the 2 groups. However, the only independent BSPM variable was the sum of ST integral decrease, averaging --2,323 +/- 1,809 microV.s for normal patients between rest and immediate cessation of exercise, compared with -3,828 +/- 2,329 microV.s for the LAD patients (p less than 0.05). Late recovery minus rest difference averaged -1,264 +/- 1,080 microV.s for normal subjects and -2,575 +/- 1,844 microV.s for LAD patients (p less than 0.01). To control for the physiologic changes of exercise, the ST integral temporal differential maps of the normal subjects were subtracted from those of the LAD patients and the sum of negative intergroup differences was assumed to reflect only ischemia. Correlation of ST integral ischemia values at immediate cessation of exercise and late recovery was high (r = 0.88); however, intertechnique correlations of the BSPM variables with quantitative angiographic scores and thallium perfusion scan scores revealed generally low r values (range 0 to 0.52). These data demonstrate that ischemic repolarization changes are detectable and quantifiable by BSPM at low levels of cardiac stress in patients with 1-vessel disease when the usual electrocardiographic criteria of myocardial ischemia are frequently absent. The data further suggest that ST integral changes reflective of myocardial ischemia persist well after the exercise recovery period and that they are complementary to, rather than substitutionary for, other indirect measures of myocardial ischemia.