Changes in cardiac gap junction expression, such as those following myocardial infarction and produced in connexin knockout mice, are associated with a predisposition to arrhythmias. The present experiments investigated the effects of heptanol, a reversible gap junction inhibitor, on isolated Langendorff-perfused rabbit hearts. The introduction and withdrawal of heptanol inhibited both pressure generation and electrical conduction. These effects were completely reversible. Possible mechanisms for these findings were investigated through measurement of the concentration dependence of heptanol's effects upon conduction velocity and repolarization duration. Low concentrations of heptanol (less than 0.3 mM) caused small but significant increases in the delay between the stimulus (delivered to the basal septum) artefact and local activation of the left ventricle, as measured from bipolar electrogram (BEG) recordings. There was a steep increase in the latency between stimulus and left-ventricular activation at concentrations of heptanol above 0.3 mM. These findings are explicable by earlier reports of heptanol actions on gap junctions in vitro and modelling studies of the effects of reduced gap junction conductance on conduction velocity. Heptanol decreased repolarization duration, measured from the activation recovery interval (ARI) of BEGs, and monophasic action potential duration at 70% repolarization (MAPD70). Heptanol also reduced left-ventricular developed pressure (LVDP), and the maximum rates of contraction and relaxation of the left ventricle; these effects were concentration dependent and reversible. However, changes in ARIs, LVDP and the maximum rates of change of pressure lacked the steep response to 0.3-1.0 mM heptanol shown by the latency. These other effects are therefore likely to be mediated by cellular targets other than gap junctions. Perfusion of hearts with heptanol was also associated with a high incidence of arrhythmias. During premature stimulation protocols arrhythmias could be induced in hearts perfused with 0.1-0.3 mM heptanol but not at higher concentrations. This suggests that there is a critical range of slowed conduction that permits the development of re-entrant arrhythmias in the normal heart, although the effects of heptanol on repolarization duration may also contribute to its pro-arrhythmic activity.