DURING a recent meeting concerning renal function one of the speakers suggested that the reason for the high metabolic rate of the mammalian kidney might be further investigated since very little energy seems to be used in forming the urine. The allusion was to a computation by Borsook and Winegarten1 showing that the thermodynamic efficiency of the kidney is about 1 per cent. However, the inference, encountered not infrequently, and not only in connexion with renal function, is based on a misunderstanding of the nature of the computation.