While the study of the physiochemical composition and structure of the interstitium on a molecular level is a large and important field in itself, the present review centered mainly on the functional consequences for the control of extracellular fluid volume. As pointed out in section I, a biological monitoring system for the total extracellular volume seems very unlikely because a major part of that volume is made up of multiple, separate, and functionally heterogeneous interstitial compartments. Even less likely is a selective volume control of each of these compartments by the nervous system. Instead, as shown by many studies cited in this review, a local autoregulation of interstitial volume is provided by automatic adjustment of the transcapillary Starling forces and lymph flow. Local vascular control of capillary pressure and surface area, of special importance in orthostasis, has been discussed in several recent reviews and was mentioned only briefly in this article. The gel-like consistency of the interstitium is attributed to glycosaminoglycans, in soft connective tissues mainly hyaluronan. However, the concept of a gel phase and a free fluid phase now seems to be replaced by the quantitatively more well-defined distribution spaces for glycosaminoglycans and plasma protein, apparently in osmotic equilibrium with each other. The protein-excluded space, determined mainly by the content of glycosaminoglycans and collagen, has been measured in vivo in many tissues, and the effect of exclusion on the oncotic buffering has been clarified. The effect of protein charge on its excluded volume and on interstitial hydraulic conductivity has been studied only in lungs and is only partly clarified. Of unknown functional importance is also the recent finding of a free interstitial hyaluronan pool with relatively rapid removal by lymph. The postulated preferential channels from capillaries to lymphatics have received little direct support. Thus the variation of plasma-to-lymph passage times for proteins may probably be ascribed to heterogeneity with respect to path length, linear velocity, and distribution volumes. Techniques for measuring interstitial fluid pressure have been refined and reevaluated, approaching some concensus on slightly negative control pressures in soft connective tissues (0 to -4 mmHg), zero, or slightly positive pressure in other tissues. Interstitial pressure-volume curves have been recorded in several tissues, and progress has been made in clarifying the dependency of interstitial compliance on glycosaminoglycan-osmotic pressure, collagen, and microfibrils.(ABSTRACT TRUNCATED AT 400 WORDS)