The purpose of this study is twofold. First, to introduce a new phenomenon - abductor vocal cord spasm, or abductor laryngospasm. Second, to reconcile a longstanding discrepancy involving the effect of lung inflation on vocal cord position. Abductor laryngospasm was induced in six anesthetized, tracheotomized dogs by the administration of continuous positive airway pressure (CPAP) after occlusion of the trachea for 60 seconds. Abductor laryngospasm was characterized by sustained vocal cord abduction with concomitant massive, continuous posterior cricoarytenoid muscle (PCA) electromyography (EMG) activity and little or no laryngeal adductor muscle EMG activity. Vocal cord opening was sustained for up to 90 seconds. In the same dogs, administration of CPAP, after hyperventilation of the animal, resulted in adductor laryngospasm. Adductor laryngospasm was characterized by steady apposition of the vocal cords, massive laryngeal adductor muscle EMG activity, and silent PCA EMG activity. Abductor laryngospasm appears to be the physiologic converse of adductor laryngospasm. The results of this study show that lung inflation can produce either vocal cord abduction or adduction, depending on whether the dog is hypoventilated or hyperventilated before administration of CPAP.