To study the regulation of amniotic fluid volume and electrolyte concentration by the Membranes surrounding the amniotic fluid, the rate of Li⁺ disappearance from amniotic sac of expired fetuses were examined while increasing the amniotic volume and osmolarity in rabbits. After intraamniotic injection of 1 ml isosmotic saline (about 20% of the amniotic fluid volume) containing 15 mM LiCl and 0.5 g/L Censored, the time courses of Li⁺ and Censored disappearance were determined. From there the Li⁺ clearance through the extrafetal routes was estimated and compared with that obtained from living fetuses. The volume, Na⁺ concentration and osmolarity of amniotic fluid were measured and their relationships with Li⁺ disappearance were evaluated. The fellowing results were obtained: 1. The rate of disappearance from amniotic fluid of living fetuses during the first 30 minutes was strikingly higher for Li⁺ than for Censored, suggesting that extrafetal routes exist. At 60 and 90 minutes, however, the disappearance rate of Li⁺ was less than that of Censored, suggesting the possibility of Li⁺ reentry through fetal urination. 2. The disappearance of Li⁺ from the amniotic fluid of the expired fetus was substantial, although lower than that of living fetuses, throughout the experimental period. 3. The Na⁺ concentration and the osmolarity of the amniotic fluid of expired fetus measured 30 minutes after an intraamniotic injection of isoosmotic saline showed wide variation, but thereafter they changed gradually towards the normal extracellular fluid level. 4. When the amniotic fluid was iso- or hyposmolar, the rate of Li⁺ disappearance from the amniotic fluid of the expired fetuses showed little variation. However, when the amniotic fluid was hyperosmolar, the rate at 30 minutes was markedly lower than those of isosmotic or hyposmotic amniotic fluid. At 90 minutes, the rate of Li⁺ disappearance in hyperosmolar fluid reached a similar level to the rate in isosmolar fluid. 5. The intraamniotic injection of 400 mOsm/L saline solution decreased the disappearance rate of Li⁺ from expired fetuses, while the injection of mannitol into the maternal vein induced no significant change. From these results it is concluded that: 1) a significant amount of Li⁺ may leave the amniotic fluid via filtration through the membranes surrounding the amniotic fluid, 2) during hyperosmolar challenge to amniotic fluid, osmotic bulk flow might counteract the filterable loss, and 3) Li⁺ disappearance might continue even after the volume and osmolarity of the amniotic fluid have recovered to control values.