1. The Na and Cl distribution spaces of freshly isolated frog muscles are 16.7 and 12.6%, respectively. These values increase to 25.6 and 23.3%, respectively, on incubation. 2. The extracellular components of both Na and Cl efflux curves are significantly smaller in freshly isolated muscles (approximately 12%) than in incubated muscles (approximately 18%). The fast exchanging A component of the extracellular space is increased more by incubation than the more slowly exchanging B component. 3. The proton magnetic resonance (p.m.r.) transverse relaxation curve for the water of freshly isolated frog muscles did not show the long, slowly relaxing tail present in curves from muscles incubated in Ringer solution. 4. When muscles were incubated in hypertonic solutions the p.m.r. transverse relaxation curves could be resolved into three components whose sizes were consistent with the components present in the sodium and chloride efflux curves. The non-exponentiality of the p.m.r. transverse relaxation curve therfore appears to arise from water in both the A and B extracellular compartments of muscle. 5. Efflux analysis indicated that the cellular Na content of both freshly isolated and incubated frog muscle is similar to that predicted by others (Lev, 1964; Armstrong & Lee, 1971; Lee & Armstrong, 1974) from measurements of intracellular Na ion activity using Na-sensitive micro-electrodes. The remainder of the tissue Na was found in the more rapidly exchanging extracellular compartments. The results of these experiments are inconsistent with the presence of a substantial fraction of bound Na in frog muscle. 6. These experiments show that muscle extracellular space is smaller in vivo than in vitro. Efflux analysis is suggested as the most accurate method of assessing extra-cellular components.