The aim of the study was to determine the in vitro rate-dependent cellular electrophysiological effects of ambasilide (10 and 20 microM/l), a new investigational antiarrhythmic agent, in canine isolated ventricular muscle and Purkinje fibers by applying the standard microelectrode technique. At the cycle length (CL) of 1000 ms, ambasilide significantly prolonged the action potential duration measured at 90% repolarization (APD(90)) in both ventricular muscle and Purkinje fibers. Ambasilide (10 microM/l) produced a more marked prolongation of APD(90) at lower stimulation frequencies in Purkinje fibers (at CL of 2000 ms = 56.0 +/- 16.1%, n = 6, versus CL of 400 ms = 15.1 +/- 3.7%, n = 6; p < 0.05), but, in 20 microM/l, this effect was considerably diminished (15.2 +/- 3.6%, n = 6, versus 7.3 +/- 5.1%, n = 6, p < 0.05). In ventricular muscle, however, both concentrations of the drug induced an almost frequency-independent lengthening of APD(90) in response to a slowing of the stimulation rate (in 20 microM/l at CL of 5000 ms = 19.0 +/- 1.5%, n = 9, versus CL of 400 ms = 16.9 +/- 1.4%, n = 9). Ambasilide induced a marked rate-dependent depression of the maximal rate of rise of the action potential upstroke (V(max)) (in 20 microM/l at CL of 300 ms = -45.1 +/- 3.9%, n = 6, versus CL of 5000 ms = -8.5 +/- 3.9%, n = 6, p < 0. 05, in ventricular muscle) and the corresponding recovery of V(max) time constant was tau = 1082.5 +/- 205.1 ms (n = 6). These data suggest that ambasilide, in addition to its Class III antiarrhythmic action, which is presumably due to its inhibitory effect on the delayed rectifier potassium current, possesses I/B type antiarrhythmic properties as a result of the inhibition of the fast sodium channels at high frequency rate with relatively fast kinetics. This latter effect may play an important role in its known less-pronounced proarrhythmic ("torsadogenic") potential.