In normal isolated, perfused canine ventricles, increased ventricular volume leads to shortening of refractoriness. To test the hypothesis that myocardium within an infarction zone is more susceptible to volume-induced changes in refractoriness than is normal myocardium, we measured strength-interval curves at low and high end-diastolic volumes at control and infarcted sites in 14 isolated, blood perfused, canine hearts with chronic (greater than 25 days) infarctions. In addition, the effect of volume load on inducing ventricular arrhythmias was studied at one to six sites in 11 hearts. Differences in refractoriness and inducibility at low (22 +/- 5 ml) and high (48 +/- 6 ml) end-diastolic volumes were compared. At control sites, volume load reduced the absolute refractory period from 178 +/- 16.5 to 175 +/- 16.7 msec (p less than 0.05), but no significant change in the relative refractory period occurred. At infarcted sites, the change in refractoriness with volume load was greater, and the absolute refractory period decreased from 171.5 +/- 21 to 160.6 +/- 26.3 msec (p less than 0.01), and the relative refractory period decreased from 180.1 +/- 22.1 to 169.9 +/- 26 msec (p = 0.05). This differential effect of volume load on refractoriness led to an increased dispersion of overall refractoriness at high volume. Infarcted sites showing the largest changes in refractoriness were characterized by patchy scars extending at least to the midmyocardium, whereas sites located within areas of transmural scar, endocardial scar, or rare microfoci of fibrosis showed no increased sensitivity to volume load. Of eight hearts in which no tachyarrhythmias were inducible during programmed electrical stimulation at low volume, four had tachyarrhythmias induced at high volume. Sites of stimulation associated with a conversion from noninducible to inducible tachyarrhythmias showed a larger degree of shortening of refractoriness (change in absolute refractory period: 24.7 +/- 16.5 vs. 3.9 +/- 6.5 msec, p less than 0.05). These data indicate that volume loading may have electrophysiologic significance and that it may be of greater functional importance under pathologic conditions.