Atrial fibrillation (AF) remains a challenge to medical therapy. Over the past several years, a variety of experimental models of AF have been developed. These have provided insights into mechanisms underlying AF and antiarrhythmic drug action against the arrhythmia. A variety of drugs effective against clinical AF, including flecainide, propafenone, procainamide, and sotalol, have been found to terminate experimental AF. All of these agents appear to act by prolonging the wavelength for atrial reentry at rapid rates, thereby increasing the size and decreasing the number of functional circuits maintaining the arrhythmia. While the ability to terminate AF is determined by refractoriness prolongation at rapid rates, refractoriness prolongation at slow rates (e.g., sinus rhythm) can prevent AF induction by premature beats. Thus, drugs with strong reverse use-dependence (like sotalol) may be much more effective in preventing than in terminating AF. Spacial heterogeneity in refractoriness is an important contributor to AF occurrence in some models, particularly vagal AF, and is reduced by some (but not all) drugs that terminate AF. New insights are being gained into mechanisms of electrical remodeling, which promotes AF maintenance when rapid atrial rates are maintained, such as during AF. This electrical remodeling may be an interesting novel target for therapy of AF. Insights into AF mechanisms obtained in experimental models of AF should help in the development of new and improved therapeutic approaches.