AbstractBull sperm extracted with 0.1% Triton X‐100 can be reactivated to full motility with 0.33 mM Mg‐ATP (sperm models). When motile sperm models are treated with 0.66 mM NiSO4, spontaneous motility is lost. During the transition to motility arrest, the beat becomes progressively more asymmetric, finally arresting at one extreme of the beat cycle. After spontaneous motility has been lost, the flagellum retains the ability to respond to mechanical stimulation. If a microprobe is used to bend the flagellum in the direction opposite to its own prevailing curvature and released, the recoil is rapid and overshoots the equilibrium position. When the same flagellum is manipulated in the opposite direction (into a tighter bend of the existing curve), the recoil is slower and does not exceed the initial bend. If a microprobe is used to carefully bend the whole flagellum into a curve, the flagellum will resume continuous beating, but only if the imposed bend is in the direction opposite the natural curvature. The reinstated beating activity (mechanical reactivation) is sustained as long as the flagellum is held by the microprobe. The rate of change of the shear angle in these mechanically reactivated, Ni2+‐inhibited sperm suggests an impaired rate of sliding on one side of the axoneme compared to similarly restrained control sperm. It appears that Ni2+has a selective inhibitory effect on the dynein arms that bend the flagellum in one direction. Furthermore, the remaining functional arms activate only when the flagellum is bent in the direction opposing their own action. © 1995 Wiley‐L