Although attempts at intravenous alimentation with protein hydrolysates were reported as early as 19391, and intravenous catheters in the superior vena cava were being used for alimentation by 19482, little attention was paid to the potential infection hazards of this therapy during the three decades in which it was being developed. Following the initial reports by Dudrick, Wilmore and their colleagues in 1968 that total parenteral nutrition (TPN) could be employed safely for long-term nutrition with apparently normal growth, development and positive nitrogen balance3’4, this technique was used widely. A review of published reports of TPN from the years 1969 and 1970 indicates that many centres were experiencing septicaemia rates of 14–23 %5–8, and one study reported a septicaemia rate of 93 %9. It was not until 1972 when Curry and Quie published their experience reporting a combined bacterial and fungal septicaemia rate of 27 % in patients receiving TPN10 that the full significance of the infectious complications of this therapy was appreciated. The gravity of the situation was emphasized a year later when Heird and his colleagues11 summarized the published experience with TPN in paediatric patients and reported a 37% septicaemia rate. By 1972, therefore, it was widely recognized that bacterial and fungal septicaemia was the major complication of TPN12.