1. The concentrations of sodium, potassium and chloride in frog and bovine lenses showed a normal intracellular ion distribution with the sum of the internal cations approximately equal to the external sum. In the cephalopod lens, however, the sum inside was much lower than that outside.2. The membrane potentials of frog, Sepiola and bovine lenses were -63, -63 and -23 mV respectively. A comparison of the electrical data with the Nernst potentials predicted from ion concentration data indicated that sodium and chloride ions as well as potassium contributed to the membrane potential in frog and bovine. In contrast, the membrane and Nernst potentials for potassium were equal in Sepiola.3. Substituting potassium for sodium in the external medium depolarized lens potentials in all three species. Estimates of the relative permeabilities of sodium, potassium and chloride were obtained by fitting the Goldman-Hodgkin-Katz equation to the potential data.4. The potassium permeability was determined directly by (42)K efflux measurements and values of 2.99, 9.83 and 3.13 (x (-8) m sec(-1)) were obtained for frog, Sepiola and bovine lenses respectively.5. The effect of raising external potassium on the efflux rate constant was determined and there was reasonable agreement between experiment and theory (Kimizuka-Koketsu) in frog and bovine lenses, but the Sepiola data indicated that the potassium permeability decreased by a factor of 2.6 when the external potassium was raised from 10 to 120 mM-K+.6. The measured specific conductances, obtained using two internal micro-electrodes, were 7.7, 15.9 and 9.9 (Sm(-2)) for frog, cephalopod and bovine lenses respectively. These data compare with computed values (Kimizuka-Koketsu theory) of 7.5, 14.1 and 17.2 (Sm(-2)).7. The effect of increasing external potassium on the conductance was also tested and there was good agreement between experiment and theory (assuming constant permeabilities) only in the amphibian lens. However, when the cephalopod data were corrected assuming a 2.6-fold decrease in P(K) for a twelvefold increase in potassium, then there was excellent agreement between experiment and theory.8. The bovine measured conductances were much lower than the theoretical values throughout the range of external potassium concentrations and several explanations were proposed to account for the discrepancies.