Esophageal electrodes have been used for recording the electromyographic (EMG) activity of the posterior cricoarytenoid muscle (PCA). To determine the specificity of this EMG technique, esophageal electrode recordings were compared with intramuscular recordings in eight anesthetized mongrel dogs. Intramuscular wire electrodes were placed in the right and left PCA, and the esophageal electrode was introduced through the nose or mouth and advanced into the upper esophagus. On direct visualization of the upper airway, the unshielded catheter electrode entered the esophagus on the right or left side. Cold block of the recurrent laryngeal nerve (RLN) ipsilateral to the esophageal electrode was associated with a marked decrease in recorded activity, whereas cold block of the contralateral RLN resulted only in a small reduction in activity. After supplemental doses of anesthesia were administered, bilateral RLN cold block essentially abolished the activity recorded with the intramuscular electrodes as well as that recorded with the esophageal electrode. Before supplemental doses of anesthesia were given, especially after vagotomy, the esophageal electrode, and in some cases the intramuscular electrodes, recorded phasic inspiratory activity not originating from the PCA. Therefore, one should be cautious in interpreting the activity recorded from esophageal electrodes as originating from the PCA, especially in conditions associated with increased respiratory efforts.