Nine tracheotomized dogs were chronically instrumented with sonomicrometer length measurement transducers and fine wire EMG electrodes in each of the four abdominal muscles: transversus abdominis (TA), internal oblique (IO), external oblique (EO) and rectus abdominis (RA). To assess the effects of anesthesia and chronic implantation of transducers on abdominal muscle resting length and length changes, muscle length measurements were made in the anesthetized dogs and then repeated in the same dogs when awake, on successive days over periods ranging from two to eight weeks. Over the duration of the study periods the awake dogs were also exposed to different experimental protocols. The experimental protocols were postural changes, expiratory threshold loading (ETL) and CO₂ rebreathing, and ETL and CO₂ rebreathing before and after reversible vagal blockade. The objectives were to determine individual abdominal muscle tonic and phasic, expiratory shortening in response to these different stimuli. We hypothesized that the individual abdominal muscles would exhibit differing amounts of tonic and phasic activity and that the internal muscle layer (TA and IO) would be recruited preferentially compared to the external muscle layer (EO and RA) in response to each of the stimuli. To assess the relative importance of vagally-mediated reflexes in the control of abdominal muscle activation, the awake dogs were exposed to ETL and progressive hypercapnia before and after reversible vagal blockade. The results of these studies demonstrate the feasibility of using chronically implanted sonomicrometer transducers for measurement of abdominal muscle length changes in awake dogs. Phasic expiratory shortening of the TA, IO and EO was increased by changes in posture from the lateral decubitis to the sitting and standing positions, by ETL and by hypercapnia. Tonic shortening of the TA and/or the IO was also present under all three conditions. There was preferential recruitment of the internal muscles under all three conditions and this preferential recruitment persisted during ETL and CO₂ rebreathing after vagal blockade. We conclude that abdominal muscle activity is mediated by vagal and chemical reflexes and that segmental reflexes from muscle proprioceptors play a role in modulating the level of individual abdominal muscle activity.