BACKGROUND Drug therapy to maintain sinus rhythm in patients with atrial fibrillation (AF) is limited by adverse effects and inadequate efficacy. There has been an increased interest in the use of class III drugs to treat AF, and several new agents have been developed, but there is little information available about mechanisms of class III drug action in AF. The present study was designed to compare the effects of two class III agents, d-sotalol and ambasilide, in dog models of experimental AF. RESULTS A previously developed dog model of sustained vagotonic AF was used to assess the ability of equal loading doses of d-sotalol and ambasilide (2 mg/kg, followed by maintenance infusions), to terminate AF and prevent its induction. At this dose, ambasilide terminated AF in 12 of 12 dogs and prevented AF induction in 10 of 12 dogs; d-sotalol terminated AF in 1 of 8 dogs (P = .001 versus ambasilide) and prevented AF induction in none of 8 dogs (P = .002). An additional dose of d-sotalol (cumulative load, 8 mg/kg) terminated AF in 7 of 8 dogs and prevented induction in 5 of 8 dogs. In an additional 6 dogs with sterile pericarditis and inducible AF, ambasilide prevented AF induction in all 6. An equal dose of d-sotalol (2 mg/kg) failed to suppress AF induction in any dog, but 8 mg/kg of d-sotalol suppressed AF induction in all. Atrial effective refractory period (AERP) was increased by both drugs. However, the effects of d-sotalol on AERP showed strong reverse use dependence, whereas those of ambasilide did not. Neither ambasilide nor d-sotalol significantly altered conduction velocity, and both increased ventricular refractoriness, with d-sotalol once again showing more reverse use dependence. Effective doses of both agents increased AERP and the wavelength for atrial reentry at rapid rates, slowing atrial activation and terminating the arrhythmia. CONCLUSIONS The class III drugs d-sotalol and ambasilide terminate AF by increasing AERP and the wavelength for reentry. Ambasilide, which has been reported to block both the rapid and slow components of the delayed rectifier (IKr and IKs), shows less reverse use dependence of effects on refractoriness than the pure IKr blocker d-sotalol, possibly explaining the greater effectiveness of ambasilide at an equal dose level. These results indicate that class III drugs can exhibit different profiles of rate-dependent action on AERP and suggest that it may be possible to develop agents that have more desirable rate-dependent profiles than pure blockers of Ikr.