It has been suggested that the mechanical condition determines the rate-limiting step of the ATPase of the myosin heads in fibers: when fibers are isometrically contracting, the ADP release kinetics are rate-limiting, but as the strain is reduced and the fibers are allowed to shorten, the ADP release kinetics accelerate and P(i) release becomes rate-limiting. We have put this idea to the test with myofibrils as a model because with these both mechanical and chemical kinetic measurements are possible. With relaxed or rapidly shortening myofibrils, P(i) release is rate-limiting and (A)M.ADP.P(i) states accumulate in the steady state [Lionne, C., et al. (1995) FEBS Lett. 364, 59]. We have now studied the kinetics of P(i) release with chemically cross-linked myofibrils that, when adequately cross-linked, appear to be a good model for isometric contraction. By using a method that is specific for free P(i) and rapid quench flow that measures the amount of (A)M.ADP.P(i) states and free P(i), we show that (A)M.ADP.P(i) states predominate which suggests that the overall ATPase is limited by P(i) release kinetics. Therefore, under our experimental conditions with myofibrils prevented from shortening, the concentration of (A)M.ADP states is low, as with rapidly shortening and relaxed myofibrils. This result is difficult to reconcile with the sensitivity of force development in fibers and myofibrils to P(i) which implies interaction of P(i) with an (A)M.ADP state. We discuss two models for accommodating the mechanical and chemical kinetics with reference to the duty cycle in skeletal muscle.