Nineteen patients with refractory but stable ventricular premature beats (VPB) received 3 medications intravenously at 24-h intervals. Lorcainide (2 mg/kg) and placebo were given double-blind in randomized sequence and the third treatment was 100 mg lidocaine, the standard reference drug. Continuous ECG recordings were made for the first 2 hours after administration to study the antiarrhythmic effect; the stability of the arrhythmia, the absence of residual and of period effects, and the interdrug differences in efficacy were statistically evaluated. Lorcainide significantly reduced the frequency of VPB during the 2-hour period, whereas the effect of lidocaine was more short-lived. The median individual peak reduction in VPB was 96% for lorcainide and 47% for lidocaine. No significant reduction in VPB was observed with placebo. Adverse effects were acceptable with either active treatment.