Several cat hindlimb muscles that exhibit differential activation (activity that is restricted to a specific region of muscle) during natural movements were studied to determine the possible roles of 1) non-uniform distribution of histochemically-identified muscle fiber-types (semitendinosus, ST; tibialis anterior, TA) or 2) mechanical heterogeneity (biceps femoris, BF; tensor fasciae latae, TFL). Using chronic recording techniques, electromyographic (EMG) activity was recorded from multiple sites of each muscle during treadmill locomotion, ear scratch, and paw shake. Standard histochemical analysis was performed on each muscle to determine fiber-type distribution. The histochemically regionalized muscles (ST and TA) were differentially active during slow locomotion; the deep regions (high in type I [SO] fibers) were active, but the superficial regions (high in type IIB [FG] fibers) were inactive. Vigorous movements (fast locomotion, ear scratch, paw shake) produced additional, synchronous activation of the superficial regions. In all movements, ST and TA activation patterns were consistent with the existence of identically timed synaptic inputs to all motoneurons within each motoneuron pool, resulting in an orderly recruitment of each whole pool. The differential activation recorded from ST and TA during slow locomotion was presumably a consequence of the non-uniform distribution of the different muscle fiber types. In contrast, differential activation of the histochemically nonregionalized, mechanically heterogeneous muscles (BF and TFL) resulted from non-synchronous activation of different muscle regions. The selective activation of BF or TFL compartments was indicative of differential synaptic inputs to, and selective recruitment of, subpopulations of the motoneuron pool, with each motoneuron subpopulation exclusively innervating physically separate regions of the muscle consistent with the regions defined by the neuromuscular territories of the major nerve branches supplying each muscle. Individual neuromuscular compartments of BF and TFL differ in their mechanical arrangements to the skeleton and in their contribution to mechanical action(s) at the hip and knee joints. Selective neural activation of mechanically distinct compartments within a mechanically heterogeneous muscle can provide highly advantageous mechanical "options" for animals that perform kinematically diverse movements. With regard to EMG recording techniques, the results of this study emphasize the need for carefully chosen EMG sampling sites and the value of knowing the muscle histochemistry, neuromuscular and musculoskeletal anatomy and possible mechanical functions prior to recording EMG.