Because stretch-induced activation may be important in generating clinically relevant arrhythmias in the heart, we delineated the ability of different types of stretches to activate ventricular tissue. Geometrically simple sheets of frog (Rana catesbeiana) ventricular tissue were mounted to allow stretches to be applied perpendicular to one edge. Every heart could be activated by a stretch pulse (n = 25), and several parameters were varied to determine their effects on mechanical activation threshold. At shorter coupling intervals, a larger stretch was needed to excite the tissue, and activation-recovery intervals were shorter, similar to previously published electrically probed strength-interval and restitution relations. Additionally, the tissue became easier to activate as the speed of the stretch increased from 0.09 to 2.6% length/ms. The increment in stretch needed for activation decreased as the baseline stretch increased from 0 to 6% length. Thus we show that mechanical activation is similar to electrical activation and that increasing uniquely mechanical parameters such as the speed of the applied stretch or baseline level of stretch can decrease the mechanical activation threshold.