In skinned skeletal muscle fibers and skinned preparations of myocardium or smooth muscle, for all conditions studied, the length traces during isotonic shortening are always found to be significantly curved. It is demonstrated that the observed curvature is not simply due to inhomogeneities on the sarcomere level in striated muscle, damaged ends of the preparations, double overlap and collision of filaments, or depletion of MgATP during the period of isotonic shortening. It is shown that the velocity of shortening can be described by an exponential function: v = vi exp (-b[SLi-SL]) with SLi: sarcomere length at the start of the release; SL: sarcomere length during isotonic shortening. Thus, instantaneous shortening velocity (v) is determined by two parameters: vi, the initial shortening velocity for SL = SLi, and b, a constant characterizing the decrease in velocity during isotonic shortening. Factors which affect isotonic shortening can do this by affecting vi, or by changing b, or both. Therefore, when analysing the effects of interventions which affect instantaneous shortening velocity, these two possibilities have to be distinguished. Since curvature of the length traces might be caused by noncross-bridge components, only vi, the initial speed of shortening, is a parameter which directly reflects kinetics of the cross-bridge cycle while instantaneous speed of shortening might also be affected by noncross-bridge factors. Analysing isotonic shortening in terms of vi and b, the effects of ionic strength, free Ca++ concentration, MgATP/MgADP ratio and temperature on unloaded isotonic shortening have been studied. For the conditions used, it can be shown that ionic strength and free Ca++ concentration only affect b without significant effect on vi, whereas MgATP/MgADP ratio and temperature affect both vi and b. This means that of these factors only MgATP/MgADP ratio and temperature affect the cross-bridge kinetics which determine the maximum speed of shortening while ionic strength and free Ca++ concentration have no such effect within the experimental error.