Observational studies have associated a low amino acid intake with adverse outcome of critical illness. Although this finding could theoretically be explained by differences in feeding tolerance related to illness severity, guidelines have recommended to administer sufficient amounts of amino acids from early onwards in the disease course. Recently, however, several high quality randomized controlled trials have not shown benefit by early amino acid supplementation and some trials even found potential harm, thus questioning this recommendation. These negative results could be related to amino acid-induced suppression of autophagy, to the inability to suppress bulk catabolism by exogenous amino acids, or to the administration of an amino acid mixture with an inappropriate composition. Currently, there is no evidence supporting administration of individual amino acid supplements during critical illness and glutamine administration may be harmful. The optimal timing, dose and composition of the amino acid mixture for critically ill patients remain unclear.