The complications of sepsis often make it necessary to suspend nutritional therapy. The origin of these is the nutritive mixture as well as the catheter connection and/or point where it has been inserted. The aim of this study is to show the need for bacteriological control of the whole process, in order to evaluate the effectiveness of the methods used, dilucidate the origin of the sepsis and establish an internal quality control. Study of different methods for bacteriological control of the nutritive mixtures, comparing them with the methods used in our Hospital, which is based on systematic culture of the mixtures, the collecting of samples after preparation and prior to perfusing the mixture through the patient, the performing of a further culture control and a bacteriological examination of the catheter. For this purpose, 28,501 nutritive mixtures were studied, corresponding to 1,782 patients. Of these, 185 samples were initially positive (0.65%) and only 59 samples showed positive cultures (0.21%). Of the 6 cases of sepsis discovered and confirmed by haemoculture, 5 corresponded to Enterobacter cloacae and 1 to Klebsiella pneumoniae.