This study investigated whether data available after the initial electrophysiologic study in patients with sustained ventricular tachyarrhythmia could identify those patients in whom serial drug testing is likely to be efficacious. One hundred six patients with inducible sustained ventricular tachyarrhythmia, whose initial study included short-term drug testing with intravenous procainamide, were evaluated. The baseline arrhythmia induced (in the absence of all antiarrhythmic drugs) was monomorphic tachycardia with a cycle length greater than 200 ms in 81 patients and ventricular flutter or fibrillation in the remaining 25 patients. After intravenous infusion of procainamide (1,250 +/- 300 mg), a ventricular tachyarrhythmia could still be induced in 80 patients during testing with up to three extrastimuli. Serial drug testing with one to four trials of oral conventional and investigational agents was then undertaken. Evaluation of 15 clinical, hemodynamic and electrophysiologic variables by stepwise logistic regression identified two independent predictors of successful response to oral antiarrhythmic drugs: 1) noninducibility of ventricular tachycardia after intravenous procainamide (p less than 0.001), and 2) left ventricular ejection fraction greater than or equal to 40% (p less than 0.05). Subgroup analysis combining each of these variables identified patients with a high, intermediate or low probability of finding a successful oral drug regimen. Patients whose arrhythmia was suppressed by intravenous procainamide had a 100% likelihood (if left ventricular ejection fraction was greater than or equal to 40%) or an 87% likelihood (if ejection fraction was less than 40%) of responding to an oral regimen.(ABSTRACT TRUNCATED AT 250 WORDS)