The comparative effects of diltiazem and verapamil in 30 patients with long-standing atrial fibrillation were evaluated in a prospective clinical trial. After a one- to two-day washout period during which drugs other than digoxin were withdrawn, patients were randomly assigned to diltiazem or verapamil treatment groups. All therapy was double blind. Both drugs were given in ascending doses as follows: days 1-6 (part I): diltiazem, 180 mg/d, or verapamil, 240 mg/d; days 7-12 (part II): diltiazem, 360 mg/d, or verapamil, 480 mg/d. Patients failing to convert to sinus rhythm after 12 days had dosage reduced to 180 mg/d of diltiazem or 240 mg/d of verapamil, and quinidine, 750 mg/d, was coadministered for another six days (part III). Medication compliance was verified by frequent measurement of serum drug concentrations. Three verapamil patients dropped out during part I due to adverse reactions (dyspnea, pulmonary congestion, skin rash, or hepatotoxicity). The higher dosage of either verapamil or diltiazem in part II was not well tolerated, and in eight patients part III had to be initiated early due to symptomatic bradycardia. Only one patient in the diltiazem group converted to sinus rhythm, whereas five converted with verapamil (two with verapamil alone, three when combined with quinidine). Thus, diltiazem and verapamil alone are unlikely to convert atrial fibrillation to sinus rhythm. The combination of verapamil and quinidine, however, is a potentially useful pharmacologic approach, having converted atrial fibrillation to sinus rhythm in nearly 50% of patients.