A 76-year-old man developed septicaemia during the infusion of stable plasma protein solution (SPPS) which was subsequently shown to be contaminated with Acinetobacter anitratus. Septicaemia persisted for four days despite change of the intravenous cannula and administration of an appropriate antibiotic. Clinical improvement occurred only when the entire intravenous line, (infusion bottle, airway needle, giving set and intravenous cannula), all of which grew Acinetobacter, was replaced. Contamination of the SPPS probably occurred in the ward via a contaminated giving set and airway needle, which had been in use for one week. This case illustrates the importance of following accepted guidelines for infection control in intravenous therapy.