The scalene is a primary respiratory muscle in humans; however, in dogs, EMG activity recorded from this muscle during inspiration was reported to derive from underlying muscles. In the present studies, origin of the activity in the medial scalene was tested in rabbits, and its distribution was compared with the muscle mechanical advantage. We assessed in anesthetized rabbits the presence of EMG activity in the scalene, sternomastoid, and parasternal intercostal muscles during quiet breathing and under resistive loading, before and after denervation of the scalene and after its additional insulation. At rest, activity was always recorded in the parasternal muscle and in the scalene bundle inserting on the third rib (medial scalene). The majority of this activity disappeared after denervation. In the bundle inserting on the fifth rib (lateral scalene), the activity was inconsistent, and a high percentage of this activity persisted after denervation but disappeared after insulation from underlying muscle layers. The sternomastoid was always silent. The fractional change in muscle length during passive inflation was then measured. The mean shortening obtained for medial and lateral scalene and parasternal intercostal was 8.0 +/- 0.7%, 5.5 +/- 0.5%, and 9.6 +/- 0.1%, respectively, of the length at functional residual capacity. Sternomastoid muscle length did not change significantly with lung inflation. We conclude that, similar to that shown in humans, respiratory activity arises from scalene muscles in rabbits. This activity is however not uniformly distributed, and a neuromechanical matching of drive is observed, so that the most effective part is also the most active.