Glycerol-treated muscle fiber bundles were subjected to three types of repetitive stretch-release cycle, and the Mg2+-ATPase activity (the rate of ADP liberation) [EC 3.6.1.3] was determined from the rate of decrease in NADH absorbance by linking the ATPase reaction with the pyruvate kinase [EC 2.7.1.40] and lactate dehydrogenase [EC 1.1.1.27] system. The fiber bundles were stretched linearly with time for 0.3 s by 2.5% of the rest length (Lo), released suddenly, then fixed at Lo for a seconds. The length change cycle was repeated (repetitive jagged stretch-release followed by return to rest length). The activity of the oscillated fibers increased with increase in the value of α, and at α=2 s, it reached a maximal level of up to 1.9 times that of static fibers. It decreased gradually with further increase in α to a value slightly larger than that of static fibers. The calculated time course of ADP liberation after the sudden release of slow stretch showed a sigmoidal increase, a decrease in the rate to almost zero, then a return to the rate of static fibers. The amount of ADP liberated at the end of the change was about 0.7 mol per mol of myosin. The fiber bundles were stretched linearly with time for β seconds by 2.5 % of Lo, then restretched immediately after a sudden release to Lo. The cycle was repeated (simple repetitive jagged stretch-release). The ATPase activity increased with increase in β and at β≈0.06 s, it reached a maximal level of up to 1.75 times that of static fibers. It decreased gradually with further increase in β to a value slightly larger than that of static fibers. The fiber bundles were stretched by 2.5% of Lo for 0.01 s, held at the plateau length, then re-stretched immediately after a sudden release to Lo. The cycle was repeated (intermittent stretch-release). The ATPase activity increased with increase in the length of the cycle (β) and at β≈ 0.05 s, it reached a maximal level of up to 1.8 times that of static fibers (the first acceleration). Then it decreased gradually with increase in β4, becoming nearly equal to that of static fibers, and increased again with further increase in β to reach a maximal level of up to 1.5 times that of static fibers at β≈1 s (the second acceleration). Finally, it decreased with further increase in β to a value slightly larger than that of static fibers. The acceleration of ATPase activity on applying a repetitive jagged stretch-release followed by return to rest length and the second acceleration on applying an intermittent stretch-release had the following common properties. Both were saturated at ΔL/Lo≈2%, showed substrate inhibition, and markedly decreased when ATP was replaced with ITP or Mg2+with Mn2+. They were suppressed by modification of the muscle fibers with PCMB. The value of α or β for the maximum acceleration decreased to one-fourth when the temperature was raised from 3−4°C to 20°C, and increased to 4–5 s when the Ca2+concentration was lowered from 10 to 0.1 μM. All these results, together with those mentioned in point 1, clearly indicate that one cycle of ATP splitting of cross-bridges is synchronously initiated by the dissociation of cross-bridges by both types of length change. On the other hand, the acceleration of ATPase activity on applying a simple repetitive jagged stretch-release and the first acceleration on applying an intermittent stretch-release had the following common properties. Both showed a bell-shaped dependence on the amplitude of the length change (ΔL/Lo) with the maximum at ΔL/Lo of 2–3%. They were almost unaffected by raising the temperature or lowering the Ca2+concentration. Furthermore, the first acceleration on intermittent stretch-release was only slightly affected by change in ATP concentration, and by replacement of ATP with ITP or Mg2+with Mn2+, while it was markedly enhanced by modification of the muscle fibers with PCMB. These results strongly suggest that the ATPase rate of one cross-bridge increases markedly at an angle greater than 90°.