In 100 patients with IHD and complex ventricular arrhythmias, programmed electrical stimulation was performed using up to three extrastimuli at sinus rhythm, and paced 100, 120 and 140 beats/min delivered from the RV apex, outflow tract or the LV with ventricular mapping to evaluate late potentials (LP) in 41 patients. Sustained monomorphic VT (SMVT) was provoked in 91% of 42 patients with a history of VT/VF, P less than 0.001, all five patients had SMVT in 24-hour ECG, P less than 0.005, and 91% of 21 patients with LV dyskinesis, P less than 0.01. After depolarizations were found in 62% of 21 patients with a history of VT, in 58% of 31 patients with inducible VT, P less than 0.01 and in five of six patients with LV dyskinesis. In patients with inducible VT, LP had a higher amplitude (105 +/- 35 vs 60 +/- 47 microV) and were more delayed (202 +/- 96 vs 133 +/- 75 msec) than in noninducible patients. In 17 patients, serial drug testing was performed after oral administration using mexilitene, disopyramide, chinidine, propafenone, sotalol, and amiodarone. If one drug was tested, the therapy efficacy was 25%, if two drugs-60%, and if three drugs-75%. In eight patients, VT was inducible in all tests, but in only one of these patients chronic antiarrhythmic therapy was not effective. We conclude that the most important predictors of VT inducibility are a history of VT or 24-hour ECG, and LV dyskinesis. Serial drug testing is efficient only when many drugs are tested, but even if VT is inducible, it does not exclude the possibility of a good clinical outcome in chronic therapy.