Teleosts and elasmobranchs faced with considerable osmotic challenges living in sea water, use compensatory mechanisms to survive the loss of water (teleosts) and urea (elasmobranchs) across epithelial surfaces. We hypothesized that the gill, with a high surface area for gas exchange must have an apical membrane of exceptionally low permeability to prevent equilibration between seawater and plasma. We isolated apical membrane vesicles from the gills of Pleuronectes americanus (winter flounder) and Squalus acanthias (dogfish shark) and demonstrated approximately sixfold enrichment of the apical marker, ADPase compared to homogenate. We also isolated basolateral membranes from shark gill (enriched 2.3-fold for Na-K-ATPase) and using stopped-flow fluorometry measured membrane permeabilities to water, urea, and NH(3). Apical membrane water permeabilities were similar between species and quite low (7.4 +/- 0.7 x 10(-4) and 6.6 +/- 0.8 x 10(-4) cm/s for shark and flounder, respectively), whereas shark basolateral membranes showed twofold higher water permeability (14 +/- 2 x 10(-4) cm/s). Permeabilities to urea and NH(3) were also low in apical membranes. Because of the much lower apical to basolateral surface area we conclude that the apical membrane represents an effective barrier. However, the values we obtained were not low enough to account for low water loss (teleosts) and urea loss (elasmobranchs) measured in vivo by others. We conclude that there are other mechanisms which permit gill epithelia to serve as effective barriers. This conclusion has implications for the function of other barrier epithelia, such as the gastric mucosa, mammalian bladder, and renal thick ascending limb.