The choice of colloid or crystalloid solutions for fluid resuscitation of critically ill patients remains controversial. Marked reduction of extracellular water is considered an important defect in shock by proponents of crystalloid fluid therapy. Large volumes of crystalloid replenish this extracellular deficit. Hypovolemia is regarded as the primary defect in shock by those favoring colloid fluid therapy. Colloidal fluids promptly restore plasma volume and reestablish hemodynamic stability with substantially lesser volumes of fluid. However, only 8% of infused water and less than 25% of infused saline are retained in the intravascular fluid compartment after 1 h. On the other hand, almost the entire volume of iso-oncotic colloid is retained with the intravascular space after 1 h. Hypertonic saline decreases intracellular fluid volume. Hyperoncotic colloid decreases both interstitial and intracellular fluid volumes as it disproportionately expands intravascular volume. The choice of fluid is not only contingent on the restoration of intravascular volume, the rapidity with which it is accomplished and the duration of its effect, but also on the adverse effects that follow fluid resuscitation. This is of greatest moment in the fluid resuscitation of patients in whom circulatory shock follows volume depletion. Crystalloid fluid repletion which requires between 2- and 4-fold as much volume as colloidal fluid is of little risk in the young, traumatically injured patient. However, in older patients, the risk of pulmonary edema is increased.