Proprioceptive information from muscle spindle sensory afferents plays a critical role in movement, yet we lack mechanistic models to tease apart how physiological and pathological at multiple scales changes alter sensorimotor control. Altered muscle spindle function is implicated in a wide range of sensorimotor impairments including dystonia, hypotonia, ataxia, and Parkinson's disease, as well as in spasticity, which affects those with stroke, cerebral palsy, spinal cord injury, and other neural injuries. But, despite decades of work, the basic mechanisms of muscle spindle sensory encoding are not well understood, and thus their role in sensorimotor disorders have not been clearly identified. Our objective is to develop a novel, mechanistic, multi-scale model of muscle spindle sensory encoding to that will allow us to test hypotheses about the role of molecular, cellular, and circuit level mechanisms on sensorimotor control in healthy and impaired humans and animals. We will build a neuromechanical muscle spindle model incorporating muscle sarcomere cross- bridge dynamics, mechanical properties of the spindle-bearing musculotendon, and biophysical membrane properties of muscle spindle afferent neurons and motor neurons. The model will be a useful platform to integrate classical and new findings of muscle spindle function spanning molecular and behavioral levels. We will identify the source of history-dependent characteristics of muscle spindle firing rates. Specifically, we will identify the mechanisms behind initial bursts, rate relaxation at constant length, and dynamic response modulation to ramps. We will dissociate the relationship of muscle spindle firing rates to kinetic (force) versus kinematic (length) variables using the same set of novel stretch perturbations applied to intact muscle in vivo in cats and rats (Aim 1), single muscle fibers from the same animals in vitro (Aim 2), and a multi-scale neuromechanical model incorporating cross-bridge dynamics, musculotendon viscoelastic properties, and spiking neuron models (Aim 3). \N_e will also use the model to interpret our existing data from animals with sensory loss.