Sodium chloride transfer across isolated frog skin is described by the well-known Koefoed Johnsen-Ussing (KU) model, the central features of which are 1) a two-step, active, inward transport of Na+, and 2) passive cotransfer of Cl-, which is coupled electrically to Na+ movement under open-circuit conditions. However, NaCl absorption by the frog skin in vivo involves active inward transport of both ions by completely independent systems. Electrical neutrality is maintained by countertransfer of H+ (exchanged for Na+) and HCO-3 (exchanged for Cl-). This behavior is called the Krogh (KR) model. The KU and KR models share some features, notably amiloride sensitivity and participation of the Na+-K+-ATPase in Na+ transport, but the differences between them are fundamental. The latter appear to be due to the use of different experimental conditions. Intact frogs are usually studied in dilute (approximatley 1 mM) external solutions, while Ringer solution is used in most work on isolated skins. The skin is virtually impermeable to Cl- in dilute external media but permeable in Ringer solution. This concentration-dependent change in PCl can explain most of the differences between KU and KR models. Regulation of blood NaCl concentration in freshwater aquatic animals requires active uptake of both Na+ and Cl-. Data on representatives of four phyla show that the KR model describes the transport behavior in all of them. Such similarities in unrelated animals suggest that the transport mechanisms evolved very early in marine ancestors of modern freshwater forms. The implications of this suggestion are considered.