1. The inotropic and electrophysiological effects of aconitine were measured in the isolated, isometrically contracting guinea-pig papillary muscle during the prearrhythmic phase of alkaloid action. 2. In muscles stimulated continually at 1 Hz, 1 mumol/l aconitine produced a positive inotropic effect that reached 38 +/- (SEM) 9% immediately before the onset of arrhythmia (n = 3). 3. If aconitine (0.5 mumol/l) was applied to non-stimulated (resting) muscles for 30 min and 1-Hz stimulation resumed thereafter, the arrhythmia occurred after 724 +/- 101 beats. Prolongation of the rest exposure to 2 h did not significantly diminish the number of prearrhythmic beats. Thus, the onset of aconitine action is critically determined by muscle activity (rather than by time), and a 30-min aconitine application to the resting muscle suffices for complete equilibration of the tissue. 4. Using the preequilibration-at-rest procedure, the positive inotropic effect of aconitine (0.25 - 4 mumol) was found (a) to be absent in the rested-state contraction, (b) to grow with both number of subsequent beats and alkaloid concentration, and (c) to reach a similar prearrhythmic maximum at all concentrations. This maximum amounted to about 1/4 of the maximum positive inotropic effect of dihydroouabain. It was not influenced by reserpine pretreatment of the guinea pig. 5. Aconitine (1 mumol/l) delayed the repolarization phase of the action potential by establishing a secondary plateau at approximately -60 mV. This effect paralleled the positive inotropic effect and, like the positive inotropic effect, was abolished by 10 mumol/l tetrodotoxin (TTX). In partially depolarized muscles ([K]0 = 24 mmol/l) aconitine (8 mumol/l) produced a TTX-sensitive increase in amplitude and rate of rise of the rested-state contraction; this indicates a voltage-dependent effect on some resting Na channels. 6. While delaying the late repolarization phase, aconitine markedly shortened the early repolarization at levels positive to -40 mV, reduced the overshoot and decreased the maximum rat of depolarization of the action potential. Slow action potentials ([K]0 = 24 mmol/l; 10 mumol/l TTX) were insensitive to aconitine. 7. We conclude that the well known property of aconitine to prolong the Na influx during the action potential leads to a positive inotropic effect, thus confirming the importance of Na influx for the regulation of myocardial contractility. The exact mechanism of an additional effect by which aconitine reduces the overshoot and shortens the plateau phase of the action potential awaits further study.