OBJECTIVE Guidelines are supposed to be helpful in clinical practice. Guidelines are also supposed to rest upon the evidence that there is. In the field of clinical nutrition the problem is that many clinical trials are not conclusive because they are underpowered and sometimes have an inferior design. RESULTS The publication of the Canadian guidelines one year ago initiated a lively debate. The Canadian guidelines used meta-analysis as a tool to review the literature. This resulted in both a sound evaluation of studies as well as some controversial recommendations. The Canadian guidelines are here put in a perspective in which the older type of guidelines are compared, and some of the points of recommendation are scrutinized. CONCLUSIONS What all guidelines agree upon is the shortage of solid knowledge, the conviction that complications related to nutritional therapy in the intensive care unit are not acceptable, and that enteral nutrition is preferable if it can be given without risk. Beyond that, many controversies remain and the need for high quality prospective studies must be emphasized. In addition, such studies must address the clinically important questions that the guidelines try to answer.