In 24 patients with stable premature ventricular contractions (PVCs) greater than or equal to 100/h, Lown class greater than or equal to 2 the relative anti-arrhythmic efficacy of flecainide 150 mg twice daily and slow-release mexiletine 360 mg twice daily was evaluated in a double-blind placebo-controlled randomized crossover study. All the patients had normal ventricular function. Criteria of efficacy were: reduction greater than or equal to 70% of PVCs or reduction greater than or equal to 50% with abolition of Lown class greater than 2 arrhythmias or suppression of non-sustained ventricular tachycardias (nSVT). Twenty-two patients completed the study protocol. The placebo phases showed comparable results and no carry over effect. The criteria of efficacy were fulfilled in 20 of the 22 patients (91%) on flecainide and in 12 of the 22 (55%) on mexiletine. The absolute reductions of PVCs, couplets and nSVT obtained on flecainide and mexiletine, in comparison to the placebo, were statistically significant (p less than 0.01 for flecainide, p less than 0.05 for mexiletine). Flecainide was superior to mexiletine in overall PVC reduction (p less than 0.05). In the 17 patients with couplets the reduction obtained with flecainide was superior to mexiletine (p less than 0.05). Both drugs were highly effective on nSVT. At steady state, the mean plasma levels of both drugs were within the range of clinical efficacy. The drugs were well tolerated and no patient withdrew because of side-effects. It was concluded that at the dosages employed flecainide was superior to mexiletine in reducing premature ventricular contractions and in abolishing couplets. The efficacy of both drugs for non-sustained ventricular tachycardias was comparable. Both drugs were highly effective by comparison with the placebo.