Energetic, kinetic and oxygen exchange experiments in the mid-1980s and early 1990s suggested that phosphate (P i ) release from actomyosin-adenosine diphosphate P i (AMÁADPÁP i ) in muscle fibres is linked to force generation and that P i release is reversible. The transition leading to the force-generating state and subsequent P i release were hypothesized to be separate, but closely linked steps. P i shortens single force-generating actomyosin interactions in an isometric optical clamp only if the conditions enable them to last 20-40 ms, enough time for P i to dissociate. Until 2003, the available crystal forms of myosin suggested a rigid coupling between movement of switch II and tilting of the lever arm to generate force, but they did not explain the reciprocal affinity myosin has for actin and nucleotides. Newer crystal forms and other structural data suggest that closing of the actin-binding cleft opens switch I (presumably decreasing nucleotide affinity). These data are all consistent with the order of events suggested before: myosinÁADPÁP i binds weakly, then strongly to actin, generating force. Then P i dissociates, possibly further increasing force or sliding.