Various experimental studies conducted in the 1970s demonstrated, at least in the physiological situation, the anabolic and/or anticatabolic properties of branched-chain amino acids (leucine, valine, isoleucine) or their ketoacid derivatives. This led to several clinical studies in the late 1970s and early 1980s that aimed to evaluate the potential benefits of BCAA supplementation in nutritional support of the critically ill. The data on burn, trauma, and sepsis are, however, far from convincing. Besides significant discrepancies in their results and the fact that most of these studies involved very small populations of patients, few of them meet the current standards of therapeutic evaluation. However, some positive results in specific studies suggest that the underlying concept may be correct but that interpretation has been faulty. Indeed, we know now that while the BCAAs possess regulatory properties on protein metabolism, leucine is by far the most potent, while isoleucine and valine are inefficient. However, in the above-mentioned studies, BCAA-supplemented nutrition very frequently supplied almost equivalent amounts of all 3 BCAAs. Moreover, several studies were performed without adequate basal nutritional support, which most probably hampered the correct metabolic utilization of these amino acids. Taken together, these factors mean that the demonstrations of BCAA efficacy were fortunate in the least. In contrast, more recently, leucine was demonstrated to positively affect protein synthesis in an experimental model of sepsis or burn. In parallel, 2 prospective controlled trials of BCAA supplementation in septic patients also demonstrated an improvement in patients' nutritional status and outcome. Thus, we should abandon the concept of BCAA-supplemented nutrition for a more promising leucine-supplemented nutrition that requires further evaluation.