The extracellular space of the amphibian lens, measured with [14C]inulin, was equivalent to 7% of the tissue water. In decapsulated lenses it was only 1·2% so that about 80% of the total extracellular space is located peripherally. The total [14C]mannitol space was larger (12%) though the capsular part was the same as that found using inulin. Mannitol has greater access to the non-capsular parts of the lens than inulin, probably reflecting its smaller molecular size. Neither inulin nor mannitol can cross the lens in significant amounts and when placed in the bathing solutions on either side are confined in the anterior and posterior spaces in the ratio 1:3. This seems to reflect the presence of a barrier which is impermeant to such molecules and may correspond to the anterior lens epithelium. Osmotic cataracts, induced with hyperosmotic mannitol, resulted in an increase of 80% in the inulin and mannitol spaces in the non-capsular regions of the lens and this change may contribute to the opacity. The Cl-space measured with 36Cl was 6% and 12·5%, 23% of the lens water after 3, 18 and 24 hr incubation respectively. This is less than estimates based on total Cl content from chemical analysis (30%-24 hr), probably reflecting a slow exchange process. The Na-space measured with 22Na was 31% of the lens water compared to 34% from chemical determinations (24 hr). The extracellular space thus contains only about 25% of the total Na and Cl in the lens, but this is probably the principal contributor to the translenticular transport of the chloride.