This study investigated the activation of cardiac tissue by 'secondary sources,' which are localized changes of the transmembrane potential (V(m)) during the application of strong extracellular electrical shocks far from the shock electrodes, in cultures of neonatal rat myocytes. Cell monolayers with small intercellular clefts (length, 45 to 270 μm; width, 20 to 70 μm [mean±SD, 54± 13 μm]; n=46) were produced using a technique of directed cell growth. Changes in V(m) relative to the action potential amplitude (ΔV(m)/APA) were measured using a fluorescent voltage-sensitive dye and a 10x10 photodiode array. Shocks with voltage gradients of 4 to 18 V/cm were applied across the clefts during either the action potential (AP) plateau or diastole. During the AP plateau, shocks induced secondary sources in the form of localized hyperpolarizations and depolarizations in the regions immediately adjacent to opposite sides of the clefts. The strength of the secondary sources, defined as the difference of ΔV(m)/APA across a cleft, increased with increasing cleft length or increasing electrical field gradient. For shocks with a gradient of 8.5 V/cm, the estimated critical cleft length necessary to reach a V(m) level corresponding to the diastolic threshold of excitation was 171 ±7 μm. Accordingly, shocks with average strength of 8.2 V/cm applied during diastole produced secondary sources that directly excited cells adjacent to the clefts when the cleft length was 196±53 μm (n=14) and that failed when the cleft length was 84 ± 23 μm (n=9, P<001). The area of earliest excitation in such cases coincided with the area of maximal depolarization induced during the plateau phase. These data suggest that small inexcitable obstacles may contribute to the V(m) changes during the application of strong extracellular electrical shocks in vivo.