1. The actions of externally and internally applied digoxin in heart ventricular muscle have been compared. 2. External application of digoxin (5 x 10(-8)--10(-7) M) had an inotropic effect, a steady level of twitch tension being reached at the end of about 3 hr. 3. Addition of an anti-digoxin antibody to the bathing solution prevented or reversed the digoxin effect, depending on the time of application. 4. The efflux of the antibody-[3H]digoxin-complex could be fitted by a single exponential with a half-time of 18 min. 5. In the absence of antibody, [3H]digoxin washout was two-compartmental with half-times of 4 and 72 min respectively. It is thought that the fast process signifies efflux from the extracellular space while the slow process reflects the washout of initially membrane bound glycoside. 6. When digoxin was applied by a cut end method, there was no effect on contractile strength. 7. The profile of radioactivity several hours following exposure to [3H]digoxin clearly indicated movement from cell to cell, the concentration of [3H]digoxin being above 10(-7) M in half the preparation at the end of 6 hr. Longitudinal diffusivity averaged 8.6 x 10(-8) cm-2 sec-1. 8. We conclude that digoxin has an inotropic effect when reaching the surface membrane of cardiac cells from the outside but is ineffective when applied from the inside.