Cardiac Arrhythmias and Risk Stratification after Myocardial infarction (CARISMA) is a prospective multicenter trial designed to document the incidence of cardiac arrhythmias after acute myocardial infarction (AMI), and to assess the predictive accuracy of various arrhythmic risk markers. In this substudy of the CARISMA trial, microvolt T-wave alternans (TWA) was assessed with specific equipment 6 weeks after AMI during bicycle exercise, atrial (A) pacing, and simultaneous ventricular and atrial (V + A) pacing in 80 patients with left ventricular ejection fraction (LVEF) <40%. The agreement between the acute test results was determined by overall proportion of concordance and the kappa statistic. Sustained TWA was observed in 24, 45, and 50% of the patients during the exercise test, A pacing, and V + A pacing, respectively. The number of indeterminate TWA was significantly lower during V + A pacing (n = 7) than exercise test (n = 34). The TWA concordance rate was 71% between exercise and V + A pacing (kappa= 0.53, P = 0.001), 79% between exercise and A pacing (kappa= 0.54, P < 0.001), and 95% between the two pacing modes (kappa= 0.89, P < 0.001). Patients with positive TWA in all tests had lower LVEF (28 +/- 7% vs 35 +/- 9%, P < 0.01) and wider QT dispersion (99 +/- 44 ms vs 67 +/- 38 ms, P < 0.01) than those with inconsistent test result. The low number of indeterminate tests and high concordance between the test results indicate that V + A pacing may provide a valuable means to assess TWA in patients who cannot complete the exercise test.