We performed 12-lead electrocardiographic monitoring in 97 patients during coronary angioplasty (PTCA) of a single vessel to correlate ischemic ST changes with clinical, angiographic and coronary hemodynamic variables and to determine the optimum lead or combination of leads for their detection. Ischemia (chest pain or ST change, group A) occurred in 79 patients (80%), but in only 15 of 23 patients (65%) with collaterals (p < 0.05). Ischemia occurred more often in left anterior descending and left circumflex PTCA than right coronary PTCA, but pain was the only manifestation more often in left circumflex and right coronary PTCA. Ischemic ST change was silent in 16% and this proportion did not differ in clinical or angiographic groups except for diabetes with 3 of 5 (60%) having silent ischemia (p < 0.05). Patients in group A (ischemia) compared to group B (no ischemia) had less severe lesions (85 ± 9 vs 91 ± 7%, p < 0.01), higher transstenotic gradients (62 ± 19 vs 53 ± 9 mm Hg, p < 0.05) and lower distal occluded pressures (24 ± 11 vs 33 ± 10 mm Hg, p < 0.01), suggesting less collateral flow. Compared with a 12-lead electrocardiogram, the best single lead for detecting ST change during PTCA in each artery had a sensitivity of 80% and this increased to 93% using the best 2 leads. The best 3 leads (V3/III/V5 for left anterior descending and III/V2/V5 for right coronary and left circumflex) increased sensitivity to 100%. In 50 of the patients, the lead showing maximum ST change during PTCA was monitored for a mean of 20 hours after PTCA. Recurrent ST change occurred in 2 patients (4%), both with ST↑, which preceded chest pain in both. Thus, ST changes during PTCA may be reliably detected by optimal choice of 3 electrocardiographic leads. The occurrence of ischemia is related principally to the vessel dilated and lesion hemodynamics that probably reflect collateral flow. Recurrent ischemia after PTCA is uncommon and may be detected by ST monitoring of a single lead. © 1990.