In this study, we examined whether the glottis is open or closed during central apnea and the effect of arterial P-o2 (Pa-o2) on this control. We hyperventilated nine 11- to 30-day-old awake nonsedated lambs via a tracheostomy for 1 min to induce central apnea. Four gas mixtures (8, 15, 21, and 30% O-2) were used. At the end of the hyperventilation period, the lambs were allowed to breathe spontaneously through intact upper airways. Using a pneumotachograph attached to a face mask, we measured airflow, and we continuously recorded electromyographic (EMG) activity of the thyroarytenoid (TA), the main glottic adductor muscle. We also studied the lateral cricoarytenoid muscle (LCA, laryngeal adductor), the posterior cricoarytenoid muscle (PCA, laryngeal abductor), the cricothyroid muscle (CT), and the diaphragm. We found that hyperventilation consistently induced hypocapnic central apnea in all nine lambs in hyperoxic conditions [30% inspiratory fraction of O-2 (FIo2)] in eight of nine lambs in normoxia or mild hypoxia (15 and 21% FIo2), and in four of seven lambs in hypoxia (8% FIo2). During baseline room air breathing, there was no glottic adductor muscle expiratory EMG activity or expiratory airflow braking. Continuous Th EMG activity began early during hyperventilation and continued throughout the central apnea, regardless of Pa(o2)irst subsequent breathing efforts were marked by expiratory flow braking and expiratory activity of the TA. The LCA and the TA demonstrated the same EMG activity pattern. Conversely, the activity of the glottic abductor (PCA), the CT, and the diaphragm (phasically active in inspiration during baseline room air breathing) disappeared during the hyperventilation period, was nil throughout central apnea, and resumed with the first inspiratory efforts. Thus, in awake nonsedated lambs, hypocapnic central apneas were accompanied by continuous glottic adduction regardless of Pa-o2. This combination could prevent alveolar gas from flowing out of the lung, thus permitting continuous alveolar blood gas exchanges during central apnea.