1. In the chronically catheterized sheep fetus between 122 and 143 days gestation the concentration of D-glucose in lung liquid was very low (usually less than 0.01 mM, the lower limit of detection of the analytical method) whereas the mean plasma concentration was 0.19 mM (S.E.M. 0.4, n = 13). 2. When the lung liquid concentration of D-glucose was raised to 1.67-5.00 mM, rapid uptake was observed until the concentration had fallen to its preceding low level. The uptake showed saturation kinetics (Vmax = 2.29-8.78 mumol/min, increasing with gestation; mean Km = 0.14 +/- 0.02 mM, n = 11, no change with gestation). This active uptake of glucose was blocked by phloridzin (10(-4) M). It was associated with a decrease in lung liquid secretion rate from which a change in net sodium flux could be inferred of an order suggesting one-to-one glucose-sodium co-transport. 3. Radiolabelled 3-O-methyl-D-glucose (3-O-meG) - a monosaccharide which is transported but not metabolized - was taken up rapidly from lung liquid and this rapid uptake was inhibited by D-glucose with 50% inhibition at 0.35 mM (+/- 0.08, n = 9). It was also inhibited by phloridzin (10(-4) M). 4. Radiolabelled 2-deoxy-D-glucose - a monosaccharide which is not a substrate for sodium-coupled transport - was taken up only very slowly from lung liquid; the rate of uptake was appropriate for passive diffusional transport and it was unaffected by the addition of D-glucose or phloridzin to lung liquid. 5. Intravenous infusion of D-glucose caused no detectable increase in the concentration of glucose in lung liquid unless phloridzin was added, when a slow increase was observed. 6. In two experiments with active transport blocked by phloridzin in lung liquid (10(-4) M), the rate of entry of labelled 3-O-meG from plasma to lung liquid was measured during intravenous infusion of this tracer for 29 and 23 h. The rates of entry were similar to the rate of efflux of the tracer from lung liquid when uptake was blocked by phloridzin or D-glucose, and similar to the rate expected for a metabolically inert tracer (i.e. it was some two orders of magnitude less than efflux from lung liquid in the absence of an inhibitor).(ABSTRACT TRUNCATED AT 400 WORDS)