We recorded airflow, tidal volume, respiratory muscle electromyogram (EMG), and chest wall configuration in eight normal newborn infants to investigate the determination of end-expiratory lung volume (EEV). The expiratory flow-volume representation was nearly linear and EMG evidence of respiratory muscle activity was absent during the latter part of expiration in both supine and upright postures, consistent with passive expiration. Occasional breaths were associated with marked retardation of expiratory airflow (braking). During unobstructed apnea, expiration proceeded to the relaxation volume (Vr) with no change in slope of the flow-volume curve. During breathing, EEV was greater than Vr observed during apnea. We calculated the difference between EEV and Vr estimated by extrapolation of the linear portion of the expiratory flow-volume curve as 14.4 +/- 5.4 ml (supine) and 11.8 +/- 2.4 ml (upright). When infants were tilted from supine to upright, expiratory duration (TE) and the expiratory time constant (tau) increased significantly. Since the increases in tau and TE offset each other, the EEV-Vr difference was similar in both postures. We propose that while braking plays a major role in the early part of expiration, as long as the final portion of expiration is passive, the dynamic maintenance of EEV above Vr depends on the relative values of tau and TE. Expiratory braking mechanisms interact with the passive mechanical properties of the respiratory system to modulate the balance between tau and TE. These mechanisms provide a neonatal breathing strategy to maintain EEV above a low Vr until the chest wall stiffens with maturity.