The Cardiac Arrhythmia Pilot Study tested the feasibility of performing a large-scale study to evaluate the effect of therapy of ventricular arrhythmias after acute myocardial infarction (AMI). Ten clinical sites identified patients with greater than or equal to 10 ventricular premature complexes (VPCs)/hour, recorded 6 to 60 days after AMI in patients with an ejection fraction greater than 0.20. Patients were randomized to receive 1 of 5 agents: encainide, flecainide, imipramine, moricizine or placebo. Successful therapy was defined as greater than or equal to 70% suppression of VPCs and greater than 90% suppression of runs of ventricular tachycardia. Randomization to a second agent occurred if a patient did not achieve adequate suppression with the initial agent. Patients initially randomized to placebo continued to receive placebo. Patients were evaluated at 3-month intervals for the next year. Of 30,763 patients screened, 10,734 (35%) were younger than 70 years old and had a qualifying AMI. A Holter recording was obtained in 3,957 patients, of whom 871 (22%) had qualifying arrhythmias and 687 were eligible. Of the 687 eligible patients, 502 (73%) were randomized. Mean age of enrolled patients was 59 years. One-half of patients were randomized within 1 month after AMI. Mean ejection fraction was 0.45, with 175 (35%) patients having an ejection fraction less than 0.40. On baseline drug-free recording, 173 (35%) patients had less than 30 VPCs/hour; 149 (30%) between 30 and 100/hour and 180 (36%) greater than or equal to 100/hour. At least 1 run of ventricular tachycardia was seen in 172 (34%) patients. Drugs taken at baseline were similar in all groups with 116 (23%) patients taking digitalis, 161 (32%) taking diuretics, 203 (41%) taking beta blockers and 200 (40%) taking calcium antagonists. Slightly more patients, 53 (51%), randomized to flecainide were taking calcium antagonists. No significant relation was noted between baseline VPC frequency and ejection fraction, but baseline VPC frequency was correlated with heart rate, arrhythmia noted before AMI and right bundle branch block. As expected, a high ejection fraction correlated with lower peak creatine kinase values, an inferior location of the infarct and fewer signs of congestive heart failure. At baseline, at least 1 adverse symptom was volunteered by 192 (39%) patients. The most common symptoms were unusual tiredness or fatigue, heart beating fast or skipping beats or headache. In this study, over 20 age- and AMI-eligible patients were identified to obtain each randomized patient. The randomization process successfully distributed baseline variables across drug groups.