Several clinical conditions, including septic shock and trauma, are accompanied by abnormalities of circulating blood volume. To maintain adequate end-organ perfusion, current guidelines on haemodynamic stabilisation recommend administration either of natural or artificial colloids or of crystalloids. However, the optimal choice, timing, and amount of fluid therapy for volume resuscitation in critically ill patients have been a controversial discussion topic for decades. Clinicians have a range of different options, including several colloid preparations, and crystalloid solutions of diverse osmolality. These compounds are differentially distributed between the intra- and extravascular as well as intra- and extracellular compartments, leading to a variety of physiological effects. Recent studies have demonstrated that the choice, timing, and amount of infusion therapy may substantially impact on clinical outcomes. In this context, it appears important to distinguish between fluid management (substitution of fluid losses) and volume therapy (replacement of intravascular volume deficit). Fluid losses can be best compensated by substitution of isotonic, electrolyte-balanced crystalloids that do not exert any negative effects on acid–base homeostasis, whereas intravascular volume depletion may be corrected by infusion of colloids until the specific transfusion trigger is met. As hyperoncotic colloids (especially albumin) may increase the risk of acute kidney injury, such plasma expanders should be avoided or only used with great caution. Dextran infusion cannot be recommended, as it is linked to a relevant risk of anaphylactoid reactions and renal failure, as well as impairment of blood coagulation. While the precise effects of gelatin on kidney function are still not fully understood, it has a high anaphylactoid potential and a limited volume effect compared with other synthetic colloids. Older hydroxyethyl starch preparations (i.e. hetastarch, hexastarch, and pentastarch) have been demonstrated to impair renal function and haemostasis. In contrast, modern, third-generation starches (tetrastarches) appear to be the most suitable synthetic colloids in critically ill patients. However, large, prospective, randomised trials are needed to explicitly clarify the optimal choice, timing, and amount of fluid therapy for volume resuscitation in critically ill patients.Adv Anaesthesiol Crit Care 2009;1(1):3–10.