Following a request from the European Commission, the Panel on Dietetic Products, Nutrition and Allergies (NDA) considered the evidence for setting Dietary Reference Values for chromium. Trivalent chromium (Cr(III)) has been postulated to be necessary for the efficacy of insulin in regulating the metabolism of carbohydrates, lipids and proteins. However, the mechanism(s) for these roles and the essential function of Cr(III) in metabolism have not been substantiated. The criteria for the essentiality of a trace element were considered. It was noted that attempts to create chromium deficiency in animal models have not produced consistent results, and that there is no evidence of essentiality of Cr(III) in animal nutrition. Evaluating the possibility of Cr(III) as an essential element for humans, the evidence from reported improvements associated with chromium supplementation in patients on total parenteral nutrition was considered to be the most convincing, but overall data do not provide sufficient information on the reversibility of the possible deficiencies and the nature of any dose-response curve in order to identify a dietary requirement for humans. The Panel concludes that no Average Requirement and no Population Reference Intake for chromium can be defined. Several studies assessed the effect of chromium supplementation on glucose and/or lipid metabolism. In the only study for which information on total chromium intake was available, there was no difference in parameters of glucose metabolism of normoglycaemic subjects between the placebo and chromium-supplemented periods. The Panel considered that there is no evidence of beneficial effects associated with chromium intake in healthy subjects. The Panel concluded that the setting of an Adequate Intake for chromium is also not appropriate.