We have reevaluated and clinically tested the current concepts of shock and resuscitation on a logical, physiological, and physical basis. We have considered the currently accepted resuscitation paradigm which is based upon the thesis that early rapid resuscitation of "lost" fluid volume is mandatory and that adequacy of resuscitation can be evaluated by central venous pressure, PAP, PAWP, pulse rate, blood pressure, and/or urine volume. Such methods also accept as natural concomitants that capillary beds are "damaged by injury"; that they "leak" salt, fluid, and albumin; and that these are expected occurrences which are injury-related. We have also examined and clinically evaluated the thesis that MAP is a primary reflector of the relationships between volume and the size of the currently available functional vascular space. (Currently available functional vascular space is mediated through the baroreceptor (stretch receptor)/neuroendocrine mechanisms.) Under this hypothesis, fluid resuscitation comprises infusion of a volume per unit time given so as to replete currently measurable fluid losses and to normalize and/or sustain MAP and the normal osmolar and oncotic relationships at the capillary/tissue interface while holding hydrostatic pressure at normal. Using burn injury as a model, we compared statistically homogeneous, randomly selected groups of burn patients who were resuscitated using a hypotonic fluid (130 mOsm/liter) alone (group R: 7 patients), hypertonic fluid (240 mOsm/liter) alone group H: 5 patients), or the hypertonic fluid containing albumin (12.5 g/liter) (group A: 7 patients). The results indicate that significantly smaller volumes of fluid were needed to resuscitate the patients in group A with a significantly more rapid normalization of physical, physiological, and biochemical parameters. We conclude that the physically and physiologically appropriate method of resuscitation, demonstrated in burn injury, comprises the use of a fluid given at a rate: (1) to maintain mean arterial and hydrostatic pressures within normal range; (2) that delivers a volume per unit time which does not exceed the capacity of the currently available functional vascular space; (3) that replaces concurrent measurable fluid losses; (4) that is hypertonic (to normalize capillary/tissue osmotic gradients); and (5) that contains colloid (to normalize capillary/tissue osmotic gradients); and (5) that contains colloid (to normalize capillary/tissue oncotic gradients). We further conclude that salt, fluid, and colloid loss into the interstitium during resuscitation frequently is due to the rate delivered and/or the physical nature of the fluid used and not to capillary bed damage outside the zone of injury.