1. In frog ventricle, adrenaline increases the size of the action potential, potentiates twitch tension, and enhances relaxation. Because tension development is directly controlled by membrane potential in frog ventricle, experiments were designed to separate the effects of adrenaline on the action potential from its effects on the development of tension.2. Comparison of the tension-voltage relations in the presence and absence of adrenaline showed that during the initial portion of the voltage clamp step, adrenaline potentiated tension, but beyond 1 sec into the clamp pulse tension was depressed.3. The time and voltage dependence of the positive inotropic effect of adrenaline during voltage clamp pulses were compatible with the kinetics of the slow inward current, which is known to be augmented by adrenaline in frog and mammalian ventricle.4. Ni(2+), which has been shown to block the slow inward current in frog ventricle, also inhibited the positive inotropic effect of adrenaline.5. The relaxant effect of adrenaline was demonstrated to be present at least as early as 600 msec after the onset of membrane depolarization. However, generally 1 sec or more of membrane depolarization was required before the relaxant effect of adrenaline predominated over its positive inotropic effect.6. In catecholamine depleted strips, the augmentation of the action potential and twitch tension in the presence of adrenaline was found to occur at a sixty-fold lower concentration than the relaxant effect as judged by suppression of KCl-induced contractures.7. Pure beta-receptor agonists reproduced completely the electromechanical effects of adrenaline on the frog ventricle. alpha-receptor agonists or antagonists had no effect on action potential or development of tension.8. Cyclic AMP and dibutyryl cyclic AMP were found to augment the frog ventricular action potential and potentiate twitch tension in reserpinized or beta-blocked frog ventricular strips. However, none of the relaxant effects of catecholamines could be reproduced by these agents alone.9. Theophylline produced changes in the action potential similar to those induced by adrenaline and mimicked both the positive inotropic and relaxant effects of the drug.10. The results suggest that the positive inotropic effects of adrenaline results mainly from changes induced in the action potential plateau. The changes are both time and voltage dependent, and if inhibited, leave the relaxant effect of adrenaline unopposed.11. The findings are consistent with a cyclic AMP-mediated mechanism of the positive inotropic effect of adrenaline. However, the role of cyclic AMP in mediating the relaxant effects of adrenaline is less clear.