The concentrations of sodium and potassium, the distribution of 36Cl, the content of water and the 86Rb influx were determined in four sets of rat lenses in 185, 245, 305 and 365 mosmol media. In each set, the potassium content was varied by substituting potassium for sodium in modified TC-199 media. The lenses were incubated in the media for 24 hr in order to approximate steady-state conditions. The results in the low potassium range (less than 21 mM K) are quite different from those in the high potassium range (21-125 mM K). In the high potassium range the lens volume is dependent on the concentration of sodium in the medium and the content of potassium is controlled by the concentration of potassium in the medium. There is no volume regulation in this range and the distribution of potassium approximates levels expected in a Donnan equilibrium. In the low potassium range, the content of electrolyte in the lens is at a higher level than would be expected from an extrapolation of the observations in the high potassium range. The equilibrium constant (EK) for potassium exceeds the estimated membrane potential (PD) so that an active process is probably responsible for the elevation of potassium content in the low potassium range. The Na+-, K+-pump is stimulated by potassium and by sodium in the medium. It is postulated that volume regulation in anisotonic media is the result of the gain or loss of solute as the lens changes from one steady-state to another with the concentration of sodium chloride in the medium acting as a trigger and determining the extent of the effect.