The antiarrhythmic efficacy of mexiletine was evaluated in 44 patients with drug-resistant ventricular tachyarrhythmias. In 33 of these patients, the efficacy of mexiletine was assessed on the basis of the results of programmed ventricular stimulation. Mexiletine did not alter the ventricular effective refractory period, the Q-Tc interval, or the methods of tachyarrhythmia induction and termination during programmed stimulation. The mean cycle length of ventricular tachycardia (VT) increased from 270 +/- 49 to 313 +/- 80 ms in 21 patients in whom VT remained inducible on mexiletine alone (p less than 0.002). Overall, VT remained inducible with methods similar to control (no drugs) inductions in 25 patients receiving mexiletine alone or in combination with a type I agent. VT induction was prevented in only 8 patients, 3 on mexiletine alone and 5 receiving mexiletine combined with another drug. Mexiletine alone (in 2 patients) or with another agent (in 3) suppressed clinical recurrence of VT in an additional 5 of 11 patients who did not undergo electrophysiologic study. These 13 patients were discharged on mexiletine alone (5 patients) or in combination with other drugs (8 patients), and remained arrhythmia-free over a mean follow-up period of 7.7 +/- 4.1 months. Adverse effects occurred in 27 of 44 patients (61%) and were gastrointestinal in 17 and/or neurologic in 22. The drug was discontinued because of adverse effects in 6 patients (14%). Thus, mexiletine has limited efficacy when used alone, but when combined with other drugs it may be useful in up to 30% of patients with drug-resistant ventricular arrhythmias. Adverse effects are relatively common.