We have tested, in guinea pig atria, how muscarinic stimulation by oxotremorine (Oxo) modifies the effects on action potential duration (APD) of two iK blockers: d-sotalol (5 microM) and ambasilide (1 microM). APD was prolonged by d-sotalol (+34.8 +/- 2.9%) and ambasilide (+54.2 +/- 5.5%). Simultaneous superfusion with Oxo 0.5 microM markedly shortened APD; this effect was larger in the presence of d-sotalol than in the presence of ambasilide (-69 +/- 2% vs. -37.4 +/- 5%; P < .05). Moreover ambasilide, but not d-sotalol, antagonized APD shortening induced by Oxo. The basis for such a difference between the two drugs was studied in patch-clamp experiments on isolated rabbit atrial and sinoatrial myocytes. Besides blocking iK (half-effective concentration: EC50 = 2 microM), ambasilide almost completely inhibited iKACh (-86 +/- 2% at 10 microM; EC50 = 1.6 microM), which was minimally affected by d-sotalol. Ambasilide 2 microM increased 10-fold the acetylcholine (ACh) required for 50% iKACh activation, and reduced maximally activated iKACh by 18.8 +/- 6.3%. When iKACh was activated through a receptor-independent mechanism, 10 microM ambasilide reduced this current by only 18.7 +/- 1.4% of its control value. Moreover, ambasilide, although not affecting the current i(f), in basal conditions, reversed its inhibition by ACh. Thus, 1) the effect of K+ channel blockers on atrial APD may be blunted by ACh; ambasilide effects are less sensitive to ACh than those of d-sotalol; 2) ambasilide, but not d-sotalol, inhibits iKACh; this probably occurs largely, although not exclusively, through muscarinic receptor antagonism. Inhibition of iKACh may account for the persistence of the effects of this drug on atrial APD despite muscarinic stimulation.