Rebreathing in the Mapleson D system with a varying fresh gas flow (FGF) was measured during an experimental set-up, which consisted of a compliance lung delivering the exhaled gases with the same, fixed concentration throughout the experiment, a commercially available Bain's system and a respirator. While changes in compliance, airway resistance an dead space hardly influenced the degree of rebreathing, changes in ventilatory pattern did so when the length of the expiratory pause was changed. Thus, with the Vt used which was not essentially greater than the capacity of the corrugated tube, a sufficiently long-lasting expiratory pause made it possible to establish non-rebreathing conditions with a FGF only slightly greater than the respiratory minute volume. Obviously, a short or no expiratory pause requires that the FGF rate be high enough to compensate for the lack of a fresh gas reserve in the corrugated tube at the start of inspiration. In such cases the length of the corrugated tube becomes of minor or no importance.