The large volume changes of some hollow viscera require a greater length range for the smooth muscle of their walls than can be accommodated by a fixed array of sliding filaments. A possible explanation is that smooth muscles adapt to length changes by forming variable numbers of contractile units in series. To test for such plasticity we examined the muscle length dependence of shortening velocity and compliance, both of which will vary directly with the number of thick filaments in series. Dog tracheal smooth muscle was studied because its cells are arrayed in long, straight, parallel bundles that span the length of the preparation. In experiments where muscle length was changed, both compliance and velocity showed a strong dependence on muscle length, varying by 1.7-fold and 2.2-fold, respectively, over a threefold range of length. The variation in isometric force was substantially less, ranging from a 1.2- to 1.3-fold in two series of experiments where length was varied by twofold to an insignificant 4% variation in a third series where a threefold length range was studied. Tetanic force was below its steady level after both stretches and releases, and increased to a steady level with 5-6 tetani at 5 min intervals. These results suggest strongly that the number of contractile units in series varies directly with the adapted muscle length. Temporary force depression after a length change would occur if the change transiently moved the filaments from their optimum overlap. The relative length independence of the adapted force is explained by the reforming of the filament lattice to produce optimum force development, with commensurate changes of velocity and compliance.