Perioperative fluid therapy is the subject of much controversy, and the results of the clinical trials investigating the effect of fluid therapy on outcome of surgery seem contradictory. The aim of this chapter is to review the evidence behind current standard fluid therapy, and to critically analyse the trials examining the effect of fluid therapy on outcome of surgery. The following conclusions are reached: current standard fluid therapy is not at all evidence-based; the evaporative loss from the abdominal cavity is highly overestimated; the non-anatomical third space loss is based on flawed methodology and most probably does not exist; the fluid volume accumulated in traumatized tissue is very small; and volume preloading of neuroaxial blockade is not effective and may cause postoperative fluid overload. The trials of 'goal-directed fluid therapy' aiming at maximal stroke volume and the trials of 'restricted intravenous fluid therapy' are also critically evaluated. The difference in results may be caused by a lax attitude towards 'standard fluid therapy' in the trials of goal-directed fluid therapy, resulting in the testing of various 'standard fluid regimens' versus 'even more fluid'. Without evidence of the existence of a non-anatomical third space loss and ineffectiveness of preloading of neuroaxial blockade, 'restricted intravenous fluid therapy' is not 'restricted', but rather avoids fluid overload by replacing only the fluid actually lost during surgery. The trials of different fluid volumes administered during outpatient surgery confirm that replacement of fluid lost improves outcome. Based on current evidence, the principles of 'restricted intravenous fluid therapy' are recommended: fluid lost should be replaced and fluid overload should be avoided.