BACKGROUND The relation between early afterdepolarizations (EADs) and changes in intracellular Ca2+ concentration ([Ca2+]i) is still unclear. In the present study, we compared spatiotemporal changes in [Ca2+]i related to EADs and delayed afterdepolarizations (DADs) induced by isoproterenol. RESULTS Isolated patch-clamped guinea pig ventricular myocytes, loaded with fluo-3 acetoxymethyl ester (fluo-3 AM), were paced at 0.1 to 2 Hz. Isoproterenol (100 nmol/L) caused alterations in both phase 2 and phase 4 of the action potential (AP), consistent with EADs and DADs, respectively. During EADs (n = 16), similar to driven APs, increases in [Ca2+]i occurred simultaneously throughout the cell, whereas during DADs (n = 25), they originated in discrete cell sites and propagated as a wave. This difference was confirmed by analysis of eight EADs and DADs coupled to the same beat. Ca2+ transient linked to EADs reached a peak relative fluorescence level (expressed as percentage of the maximal level reached during the last stimulated beat) that was always higher than that reached during the DADs (77 +/- 3% versus 64 +/- 2%, P < .001). Spatial heterogeneity of Ca2+ transients was assessed by the maximal time interval between peaks monitored in different cell regions; this time lag was always greater during DADs than during EADs (290 versus 40 milliseconds, P = .006). CONCLUSIONS The present study had two main findings. First, even very modest notches occurring during the plateau of the AP may be accompanied by a marked secondary increase in [Ca2+]i. Second, these Ca2+ transients occurring during EADs are synchronous throughout the cell and differ significantly from those observed under identical conditions during DADs.