We randomized 61 patients with paroxysmal atrial fibrillation (AF) ( < 48 hours from onset) to either sotalol or quinidine treatment. Conversion of rhythm was recorded by Holter monitoring. The starting 80 mg dose of sotalol was repeated at 2, 6, and 10 hours if AF persisted (heart rate > 80 beats/min), and if systolic blood was > or = 120 mm Hg. In the quinidine group, if heart rate > 100 beats/min, it was decreased with intravenous digoxin, whereafter 200 mg of oral quinidine sulfate was given maximally 3 times, each dose 2 hours apart. Conversion of AF to sinus rhythm occurred in 17 or 33 patients (52%) taking sotalol, and in 24 of 28 patients (86%) taking quinidine (p < 0.0001). Electric cardioversion was necessary in 39% of the former and in 14% of the latter group. The mean delay from first trial drug to sinus rhythm with the trial medication was 10.2 +/- 7.6 hours in the sotalol group and 4.0 +/- 2.9 hours in the quinidine group (p < 0.01). Treatment was discontinued in 16 patients taking sotalol (48%) because of asymptomatic bradycardia or hypotension, and in 20 taking quinidine (71%) because of rhythm conversion. Asymptomatic wide complex tachycardia (QRS > 0.12 second) was found in 13% and 27% of patients taking sotalol and quinidine, respectively. The longest RR intervals were 6.4 and 3.8 seconds in the sotalol and quinidine groups, respectively. Oral sotalol did not appear as effective as quinidine sulfate treatment in conversion of paroxysmal AF.(ABSTRACT TRUNCATED AT 250 WORDS)