Isolated, superfused with [Ca2+]0 = 2.5-3 mM (T = 37 degrees C) and voltage clamped ventricular myocytes of guinea-pig hearts were stimulated by pulses from a holding potential -40 mV to +5 mV (duration 300 ms). They activated L-type Ca2+ current and a biphasic contractile response: a phasic component of amplitude approximately 7% of resting cell length (duration approximately 150 ms) and a tonic component of amplitude approximately 3% of resting cell length. The phasic component was inhibited by 10(-6) M thapsigargin (Tg). Pulses from -40 mV to +5 mV stimulated a similar bi-phasic contractile response in 74% of cells (n = 126) superfused from the beginning of a 30 s period of rest with 5-10 mM Ni2+ which blocked the Ca2+ current and Na+ -Ca2+ exchanger (Ni2+-contractions). Thus, the Ni2+-contractions could be activated only by intracellular Ca2+ release. The phasic component of control contractions showed the bell-shaped voltage relation at [Ca]0 = 2 mM and sigmoid relation at [Ca]0 = 3 mM. The phasic component o Ni2+-contractions showed a sigmoid relation at voltages from -40mV to +100 mV and could not be activated at [Ca]0 = 2 mM. It was inhibited by 20 microM nifedipine, a blocker of dihydropyridine receptors, even when activated by the pulses to +70 mV, during which the Ca2+ current does not flow. We proved that nifedipine does not affect Na+-Ca2+ exchange. The phasic component of Ni2+-contractions was also inhibited by 2 nM indolizinesulphone SR33557, another dihydropyridine receptor blocker, which halved the phasic component of contractions in sontrol cells without any significant effect on the Ca2+ current. Stimulation did not activate contraction in any of 19 cells in which 20 microM nifedipine was superfused from the beginning of 30 s rest instead of 5 mM Ni2+. These cells were depolarized to +5 mV over the rest period in order to prevent intracellular Ca2+ loss by Na+ -Ca2+ exchange. Residual Ca2+ currents were much stronger in cells superfused with nifedipine than residual currents in cells superfused with Ni2+ (hardly visible in the records). Our results suggest that a vestigial remnant of a voltage-sensing mechanism similar to that in the skeletal muscle may trigger the Ca2+ release from the SR of cardiac myocytes under specific experimental conditions. In normal cells it may be complementary to calcium induced calcium release (CICR).